---
_id: '18118'
abstract:
- lang: eng
text: We introduce a new framework for studying meta-learning methods using PAC-Bayesian
theory. Its main advantage over previous work is that it allows for more flexibility
in how the transfer of knowledge between tasks is realized. For previous approaches,
this could only happen indirectly, by means of learning prior distributions over
models. In contrast, the new generalization bounds that we prove express the process
of meta-learning much more directly as learning the learning algorithm that should
be used for future tasks. The flexibility of our framework makes it suitable to
analyze a wide range of meta-learning mechanisms and even design new mechanisms.
Other than our theoretical contributions we also show empirically that our framework
improves the prediction quality in practical meta-learning mechanisms.
alternative_title:
- PMLR
article_processing_charge: No
arxiv: 1
author:
- first_name: Hossein
full_name: Zakerinia, Hossein
id: 653bd8b6-f394-11eb-9cf6-c0bbf6cd78d4
last_name: Zakerinia
- first_name: Amin
full_name: Behjati, Amin
last_name: Behjati
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Zakerinia H, Behjati A, Lampert C. More flexible PAC-Bayesian meta-learning
by learning learning algorithms. In: Proceedings of the 41st International
Conference on Machine Learning. Vol 235. ML Research Press; 2024:58122-58139.'
apa: 'Zakerinia, H., Behjati, A., & Lampert, C. (2024). More flexible PAC-Bayesian
meta-learning by learning learning algorithms. In Proceedings of the 41st International
Conference on Machine Learning (Vol. 235, pp. 58122–58139). Vienna, Austria:
ML Research Press.'
chicago: Zakerinia, Hossein, Amin Behjati, and Christoph Lampert. “More Flexible
PAC-Bayesian Meta-Learning by Learning Learning Algorithms.” In Proceedings
of the 41st International Conference on Machine Learning, 235:58122–39. ML
Research Press, 2024.
ieee: H. Zakerinia, A. Behjati, and C. Lampert, “More flexible PAC-Bayesian meta-learning
by learning learning algorithms,” in Proceedings of the 41st International
Conference on Machine Learning, Vienna, Austria, 2024, vol. 235, pp. 58122–58139.
ista: 'Zakerinia H, Behjati A, Lampert C. 2024. More flexible PAC-Bayesian meta-learning
by learning learning algorithms. Proceedings of the 41st International Conference
on Machine Learning. ICML: International Conference on Machine Learning, PMLR,
vol. 235, 58122–58139.'
mla: Zakerinia, Hossein, et al. “More Flexible PAC-Bayesian Meta-Learning by Learning
Learning Algorithms.” Proceedings of the 41st International Conference on Machine
Learning, vol. 235, ML Research Press, 2024, pp. 58122–39.
short: H. Zakerinia, A. Behjati, C. Lampert, in:, Proceedings of the 41st International
Conference on Machine Learning, ML Research Press, 2024, pp. 58122–58139.
conference:
end_date: 2024-07-27
location: Vienna, Austria
name: 'ICML: International Conference on Machine Learning'
start_date: 2024-07-21
corr_author: '1'
date_created: 2024-09-22T22:01:45Z
date_published: 2024-09-01T00:00:00Z
date_updated: 2024-10-01T09:30:03Z
day: '01'
department:
- _id: ChLa
external_id:
arxiv:
- '2402.04054'
intvolume: ' 235'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: ' https://doi.org/10.48550/arXiv.2402.04054'
month: '09'
oa: 1
oa_version: Published Version
page: 58122-58139
publication: Proceedings of the 41st International Conference on Machine Learning
publication_identifier:
eissn:
- 2640-3498
publication_status: published
publisher: ML Research Press
quality_controlled: '1'
scopus_import: '1'
status: public
title: More flexible PAC-Bayesian meta-learning by learning learning algorithms
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 235
year: '2024'
...
---
DOAJ_listed: '1'
OA_place: publisher
OA_type: gold
_id: '18856'
abstract:
- lang: eng
text: This research is aimed to solve the tweet/user geolocation prediction task
and provide a flexible methodology for the geo-tagging of textual big data. The
suggested approach implements neural networks for natural language processing
(NLP) to estimate the location as coordinate pairs (longitude, latitude) and two-dimensional
Gaussian Mixture Models (GMMs). The scope of proposed models has been finetuned
on a Twitter dataset using pretrained Bidirectional Encoder Representations from
Transformers (BERT) as base models. Performance metrics show a median error of
fewer than 30 km on a worldwide-level, and fewer than 15 km on the US-level datasets
for the models trained and evaluated on text features of tweets' content and metadata
context. Our source code and data are available at https://github.com/K4TEL/geo-twitter.git.
acknowledgement: The authors acknowledge the Institute of Science and Technology (ISTA)
for their material support and for granting access to the Twitter database archive,
which was essential for the research.
article_processing_charge: Yes
article_type: original
author:
- first_name: Kateryna
full_name: Lutsai, Kateryna
last_name: Lutsai
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: Lutsai K, Lampert C. Predicting the geolocation of tweets using transformer
models on customized data. Journal of Spatial Information Science. 2024;(29):69-99.
doi:10.5311/JOSIS.2024.29.295
apa: Lutsai, K., & Lampert, C. (2024). Predicting the geolocation of tweets
using transformer models on customized data. Journal of Spatial Information
Science. University of Maine. https://doi.org/10.5311/JOSIS.2024.29.295
chicago: Lutsai, Kateryna, and Christoph Lampert. “Predicting the Geolocation of
Tweets Using Transformer Models on Customized Data.” Journal of Spatial Information
Science. University of Maine, 2024. https://doi.org/10.5311/JOSIS.2024.29.295.
ieee: K. Lutsai and C. Lampert, “Predicting the geolocation of tweets using transformer
models on customized data,” Journal of Spatial Information Science, no.
29. University of Maine, pp. 69–99, 2024.
ista: Lutsai K, Lampert C. 2024. Predicting the geolocation of tweets using transformer
models on customized data. Journal of Spatial Information Science. (29), 69–99.
mla: Lutsai, Kateryna, and Christoph Lampert. “Predicting the Geolocation of Tweets
Using Transformer Models on Customized Data.” Journal of Spatial Information
Science, no. 29, University of Maine, 2024, pp. 69–99, doi:10.5311/JOSIS.2024.29.295.
short: K. Lutsai, C. Lampert, Journal of Spatial Information Science (2024) 69–99.
corr_author: '1'
date_created: 2025-01-19T23:01:53Z
date_published: 2024-12-26T00:00:00Z
date_updated: 2025-06-05T13:47:12Z
day: '26'
ddc:
- '500'
department:
- _id: ChLa
doi: 10.5311/JOSIS.2024.29.295
file:
- access_level: open_access
checksum: b82413f00398ffb5168e8e747571a98d
content_type: application/pdf
creator: dernst
date_created: 2025-01-20T08:41:10Z
date_updated: 2025-01-20T08:41:10Z
file_id: '18857'
file_name: 2024_JourSpatialInfoScience_Lutsai.pdf
file_size: 7250655
relation: main_file
success: 1
file_date_updated: 2025-01-20T08:41:10Z
has_accepted_license: '1'
issue: '29'
language:
- iso: eng
license: https://creativecommons.org/licenses/by/3.0/
month: '12'
oa: 1
oa_version: Published Version
page: 69-99
publication: Journal of Spatial Information Science
publication_identifier:
eissn:
- 1948-660X
publication_status: published
publisher: University of Maine
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://github.com/K4TEL/geo-twitter.git
scopus_import: '1'
status: public
title: Predicting the geolocation of tweets using transformer models on customized
data
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/3.0/legalcode
name: Creative Commons Attribution 3.0 Unported (CC BY 3.0)
short: CC BY (3.0)
type: journal_article
user_id: 68b8ca59-c5b3-11ee-8790-cd641c68093d
year: '2024'
...
---
OA_place: publisher
OA_type: gold
_id: '18875'
abstract:
- lang: eng
text: Current state-of-the-art methods for differentially private model training
are based on matrix factorization techniques. However, these methods suffer from
high computational overhead because they require numerically solving a demanding
optimization problem to determine an approximately optimal factorization prior
to the actual model training. In this work, we present a new matrix factorization
approach, BSR, which overcomes this computational bottleneck. By exploiting properties
of the standard matrix square root, BSR allows to efficiently handle also large-scale
problems. For the key scenario of stochastic gradient descent with momentum and
weight decay, we even derive analytical expressions for BSR that render the computational
overhead negligible. We prove bounds on the approximation quality that hold both
in the centralized and in the federated learning setting. Our numerical experiments
demonstrate that models trained using BSR perform on par with the best existing
methods, while completely avoiding their computational overhead.
alternative_title:
- Advances in Neural Information Processing Systems
article_processing_charge: No
arxiv: 1
author:
- first_name: Nikita
full_name: Kalinin, Nikita
id: 4b14526e-14d2-11ed-ba64-c14c9553d137
last_name: Kalinin
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Kalinin N, Lampert C. Banded square root matrix factorization for differentially
private model training. In: 38th Annual Conference on Neural Information Processing
Systems. Vol 37. Neural Information Processing Systems Foundation; 2024.'
apa: 'Kalinin, N., & Lampert, C. (2024). Banded square root matrix factorization
for differentially private model training. In 38th Annual Conference on Neural
Information Processing Systems (Vol. 37). Vancouver, Canada: Neural Information
Processing Systems Foundation.'
chicago: Kalinin, Nikita, and Christoph Lampert. “Banded Square Root Matrix Factorization
for Differentially Private Model Training.” In 38th Annual Conference on Neural
Information Processing Systems, Vol. 37. Neural Information Processing Systems
Foundation, 2024.
ieee: N. Kalinin and C. Lampert, “Banded square root matrix factorization for differentially
private model training,” in 38th Annual Conference on Neural Information Processing
Systems, Vancouver, Canada, 2024, vol. 37.
ista: 'Kalinin N, Lampert C. 2024. Banded square root matrix factorization for differentially
private model training. 38th Annual Conference on Neural Information Processing
Systems. NeurIPS: Neural Information Processing Systems, Advances in Neural Information
Processing Systems, vol. 37.'
mla: Kalinin, Nikita, and Christoph Lampert. “Banded Square Root Matrix Factorization
for Differentially Private Model Training.” 38th Annual Conference on Neural
Information Processing Systems, vol. 37, Neural Information Processing Systems
Foundation, 2024.
short: N. Kalinin, C. Lampert, in:, 38th Annual Conference on Neural Information
Processing Systems, Neural Information Processing Systems Foundation, 2024.
conference:
end_date: 2024-12-16
location: Vancouver, Canada
name: 'NeurIPS: Neural Information Processing Systems'
start_date: 2024-12-16
corr_author: '1'
date_created: 2025-01-24T17:58:16Z
date_published: 2024-12-01T00:00:00Z
date_updated: 2025-05-14T11:34:20Z
day: '01'
ddc:
- '000'
department:
- _id: GradSch
- _id: ChLa
external_id:
arxiv:
- '2405.13763'
file:
- access_level: open_access
checksum: a216cab8eddc1fe7840aede0e2c0d41e
content_type: application/pdf
creator: dernst
date_created: 2025-01-27T09:52:15Z
date_updated: 2025-01-27T09:52:15Z
file_id: '18888'
file_name: 2024_NeurIPS_Nikita.pdf
file_size: 1144656
relation: main_file
success: 1
file_date_updated: 2025-01-27T09:52:15Z
has_accepted_license: '1'
intvolume: ' 37'
language:
- iso: eng
license: https://creativecommons.org/licenses/by/4.0/
month: '12'
oa: 1
oa_version: Published Version
publication: 38th Annual Conference on Neural Information Processing Systems
publication_identifier:
eissn:
- 1049-5258
publication_status: published
publisher: Neural Information Processing Systems Foundation
quality_controlled: '1'
scopus_import: '1'
status: public
title: Banded square root matrix factorization for differentially private model training
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 37
year: '2024'
...
---
OA_place: publisher
OA_type: gold
_id: '18891'
abstract:
- lang: eng
text: "Deep neural networks (DNNs) exhibit a surprising structure in their final
layer\r\nknown as neural collapse (NC), and a growing body of works has currently
investigated the propagation of neural collapse to earlier layers of DNNs – a
phenomenon\r\ncalled deep neural collapse (DNC). However, existing theoretical
results are restricted to special cases: linear models, only two layers or binary
classification.\r\nIn contrast, we focus on non-linear models of arbitrary depth
in multi-class classification and reveal a surprising qualitative shift. As soon
as we go beyond two\r\nlayers or two classes, DNC stops being optimal for the
deep unconstrained features\r\nmodel (DUFM) – the standard theoretical framework
for the analysis of collapse.\r\nThe main culprit is a low-rank bias of multi-layer
regularization schemes: this bias\r\nleads to optimal solutions of even lower
rank than the neural collapse. We support\r\nour theoretical findings with experiments
on both DUFM and real data, which show\r\nthe emergence of the low-rank structure
in the solution found by gradient descent."
acknowledged_ssus:
- _id: ScienComp
acknowledgement: Marco Mondelli is partially supported by the 2019 Lopez-Loreta prize.
This research was supported by the Scientific Service Units (SSU) of ISTA through
resources provided by Scientific Computing (SciComp).
alternative_title:
- Advances in Neural Information Processing Systems
article_processing_charge: No
arxiv: 1
author:
- first_name: Peter
full_name: Súkeník, Peter
id: d64d6a8d-eb8e-11eb-b029-96fd216dec3c
last_name: Súkeník
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
- first_name: Marco
full_name: Mondelli, Marco
id: 27EB676C-8706-11E9-9510-7717E6697425
last_name: Mondelli
orcid: 0000-0002-3242-7020
citation:
ama: 'Súkeník P, Lampert C, Mondelli M. Neural collapse versus low-rank bias: Is
deep neural collapse really optimal? In: 38th Annual Conference on Neural Information
Processing Systems. Vol 37. Neural Information Processing Systems Foundation;
2024.'
apa: 'Súkeník, P., Lampert, C., & Mondelli, M. (2024). Neural collapse versus
low-rank bias: Is deep neural collapse really optimal? In 38th Annual Conference
on Neural Information Processing Systems (Vol. 37). Vancouver, Canada: Neural
Information Processing Systems Foundation.'
chicago: 'Súkeník, Peter, Christoph Lampert, and Marco Mondelli. “Neural Collapse
versus Low-Rank Bias: Is Deep Neural Collapse Really Optimal?” In 38th Annual
Conference on Neural Information Processing Systems, Vol. 37. Neural Information
Processing Systems Foundation, 2024.'
ieee: 'P. Súkeník, C. Lampert, and M. Mondelli, “Neural collapse versus low-rank
bias: Is deep neural collapse really optimal?,” in 38th Annual Conference on
Neural Information Processing Systems, Vancouver, Canada, 2024, vol. 37.'
ista: 'Súkeník P, Lampert C, Mondelli M. 2024. Neural collapse versus low-rank bias:
Is deep neural collapse really optimal? 38th Annual Conference on Neural Information
Processing Systems. NeurIPS: Neural Information Processing Systems, Advances in
Neural Information Processing Systems, vol. 37.'
mla: 'Súkeník, Peter, et al. “Neural Collapse versus Low-Rank Bias: Is Deep Neural
Collapse Really Optimal?” 38th Annual Conference on Neural Information Processing
Systems, vol. 37, Neural Information Processing Systems Foundation, 2024.'
short: P. Súkeník, C. Lampert, M. Mondelli, in:, 38th Annual Conference on Neural
Information Processing Systems, Neural Information Processing Systems Foundation,
2024.
conference:
end_date: 2024-12-16
location: Vancouver, Canada
name: 'NeurIPS: Neural Information Processing Systems'
start_date: 2024-12-16
corr_author: '1'
date_created: 2025-01-27T11:15:18Z
date_published: 2024-12-01T00:00:00Z
date_updated: 2025-06-04T07:19:21Z
day: '01'
ddc:
- '000'
department:
- _id: GradSch
- _id: MaMo
- _id: ChLa
external_id:
arxiv:
- '2405.14468'
file:
- access_level: open_access
checksum: b7b79f1ea3ac1e9e11b3d91faaeb0780
content_type: application/pdf
creator: dernst
date_created: 2025-02-04T08:11:25Z
date_updated: 2025-02-04T08:11:25Z
file_id: '18989'
file_name: 2024_NeurIPS_Sukenik.pdf
file_size: 1784118
relation: main_file
success: 1
file_date_updated: 2025-02-04T08:11:25Z
has_accepted_license: '1'
intvolume: ' 37'
language:
- iso: eng
month: '12'
oa: 1
oa_version: Published Version
project:
- _id: 059876FA-7A3F-11EA-A408-12923DDC885E
name: Prix Lopez-Loretta 2019 - Marco Mondelli
publication: 38th Annual Conference on Neural Information Processing Systems
publication_status: published
publisher: Neural Information Processing Systems Foundation
quality_controlled: '1'
status: public
title: 'Neural collapse versus low-rank bias: Is deep neural collapse really optimal?'
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 37
year: '2024'
...
---
OA_place: repository
OA_type: green
_id: '19063'
abstract:
- lang: eng
text: "Instruction-tuned Large Language Models (LLMs) show impressive results in
numerous practical applications, but they lack essential safety features that
are common in other areas of computer science, particularly an explicit separation
of instructions and data. This makes them vulnerable to manipulations such as
indirect prompt injections and generally unsuitable for safety-critical tasks.
Surprisingly, there is currently no established definition or benchmark to quantify
this phenomenon. In this work, we close this gap by introducing a formal measure
for instruction-data separation and an empirical variant that is calculable from
a model's outputs. We also present a new dataset, SEP, that allows estimating
the measure for real-world models. Our results on various LLMs show that the problem
of instruction-data separation is real: all models fail to achieve high separation,
and canonical mitigation techniques, such as prompt engineering and fine-tuning,
either fail to substantially improve separation or reduce model utility. The source
code and SEP dataset are openly accessible at https://github.com/egozverev/Shold-It-Be-Executed-Or-Processed.\r\n"
acknowledged_ssus:
- _id: ScienComp
acknowledgement: The authors would like to sincerely thank Juan Rocamonde for valuable
feedback to our manuscript. We acknowledge the support from the Scientific Service
Units (SSU) of ISTA through resources provided by Scientific Computing (SciComp).
We thank Dan Alistarh for providing us with computational resources. This work was
partially funded by the German Federal Ministry of Education and Research (BMBF)
under the grant AIgenCY (16KIS2012) and ELSA – European Lighthouse on Secure and
Safe AI funded by the European Union under grant agreement No. 101070617. Views
and opinions expressed are however those of the authors only and do not necessarily
reflect those of the European Union or European Commission. Neither the European
Union nor the European Commission can be held responsible for them.
article_number: '2403.06833'
article_processing_charge: No
arxiv: 1
author:
- first_name: Egor
full_name: Zverev, Egor
id: 05162b19-1340-11ed-8f02-fa94e0e8c3bc
last_name: Zverev
- first_name: Sahar
full_name: Abdelnabi, Sahar
last_name: Abdelnabi
- first_name: Soroush
full_name: Tabesh, Soroush
id: 06000900-6068-11ef-8d61-c2472ef2e752
last_name: Tabesh
orcid: 0009-0003-4119-6281
- first_name: Mario
full_name: Fritz, Mario
last_name: Fritz
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: Zverev E, Abdelnabi S, Tabesh S, Fritz M, Lampert C. Can LLMs separate instructions
from data? And what do we even mean by that? arXiv. 2024. doi:10.48550/arXiv.2403.06833
apa: Zverev, E., Abdelnabi, S., Tabesh, S., Fritz, M., & Lampert, C. (2024).
Can LLMs separate instructions from data? And what do we even mean by that? arXiv.
https://doi.org/10.48550/arXiv.2403.06833
chicago: Zverev, Egor, Sahar Abdelnabi, Soroush Tabesh, Mario Fritz, and Christoph
Lampert. “Can LLMs Separate Instructions from Data? And What Do We Even Mean by
That?” ArXiv, 2024. https://doi.org/10.48550/arXiv.2403.06833.
ieee: E. Zverev, S. Abdelnabi, S. Tabesh, M. Fritz, and C. Lampert, “Can LLMs separate
instructions from data? And what do we even mean by that?,” arXiv. 2024.
ista: Zverev E, Abdelnabi S, Tabesh S, Fritz M, Lampert C. 2024. Can LLMs separate
instructions from data? And what do we even mean by that? arXiv, 2403.06833.
mla: Zverev, Egor, et al. “Can LLMs Separate Instructions from Data? And What Do
We Even Mean by That?” ArXiv, 2403.06833, 2024, doi:10.48550/arXiv.2403.06833.
short: E. Zverev, S. Abdelnabi, S. Tabesh, M. Fritz, C. Lampert, ArXiv (2024).
corr_author: '1'
date_created: 2025-02-20T10:13:42Z
date_published: 2024-03-01T00:00:00Z
date_updated: 2025-02-24T12:52:23Z
day: '01'
ddc:
- '000'
department:
- _id: GradSch
- _id: ChLa
doi: 10.48550/arXiv.2403.06833
external_id:
arxiv:
- '2403.06833'
file:
- access_level: open_access
checksum: 35eb43968684b87be59144603ef10af0
content_type: application/pdf
creator: ezverev
date_created: 2025-02-20T10:11:45Z
date_updated: 2025-02-20T10:11:45Z
file_id: '19064'
file_name: 2403.06833v3.pdf
file_size: 530972
relation: main_file
success: 1
file_date_updated: 2025-02-20T10:11:45Z
has_accepted_license: '1'
language:
- iso: eng
license: https://creativecommons.org/licenses/by-sa/4.0/
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2403.06833
month: '03'
oa: 1
oa_version: Preprint
publication: arXiv
publication_status: published
related_material:
link:
- relation: software
url: ' https://github.com/egozverev/Shold-It-Be-Executed-Or-Processed'
status: public
title: Can LLMs separate instructions from data? And what do we even mean by that?
tmp:
image: /images/cc_by_sa.png
legal_code_url: https://creativecommons.org/licenses/by-sa/4.0/legalcode
name: Creative Commons Attribution-ShareAlike 4.0 International Public License (CC
BY-SA 4.0)
short: CC BY-SA (4.0)
type: preprint
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2024'
...
---
OA_place: publisher
OA_type: diamond
_id: '19408'
abstract:
- lang: eng
text: 'Continual learning is a subfield of machine learning, which aims to allow
machine learning models to continuously learn on new data, by accumulating knowledge
without forgetting what was learned in the past. In this work, we take a step
back, and ask: "Why should one care about continual learning in the first place?".
We set the stage by examining recent continual learning papers published at four
major machine learning conferences, and show that memory-constrained settings
dominate the field. Then, we discuss five open problems in machine learning, and
even though they might seem unrelated to continual learning at first sight, we
show that continual learning will inevitably be part of their solution. These
problems are model editing, personalization and specialization, on-device learning,
faster (re-)training and reinforcement learning. Finally, by comparing the desiderata
from these unsolved problems and the current assumptions in continual learning,
we highlight and discuss four future directions for continual learning research.
We hope that this work offers an interesting perspective on the future of continual
learning, while displaying its potential value and the paths we have to pursue
in order to make it successful. This work is the result of the many discussions
the authors had at the Dagstuhl seminar on Deep Continual Learning, in March 2023.'
alternative_title:
- TMLR
article_processing_charge: No
article_type: original
arxiv: 1
author:
- first_name: Eli
full_name: Verwimp, Eli
last_name: Verwimp
- first_name: Rahaf
full_name: Aljundi, Rahaf
last_name: Aljundi
- first_name: Shai
full_name: Ben-David, Shai
last_name: Ben-David
- first_name: Matthias
full_name: Bethge, Matthias
last_name: Bethge
- first_name: Andrea
full_name: Cossu, Andrea
last_name: Cossu
- first_name: Alexander
full_name: Gepperth, Alexander
last_name: Gepperth
- first_name: Tyler L.
full_name: Hayes, Tyler L.
last_name: Hayes
- first_name: Eyke
full_name: Hüllermeier, Eyke
last_name: Hüllermeier
- first_name: Christopher
full_name: Kanan, Christopher
last_name: Kanan
- first_name: Dhireesha
full_name: Kudithipudi, Dhireesha
last_name: Kudithipudi
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
- first_name: Martin
full_name: Mundt, Martin
last_name: Mundt
- first_name: Razvan
full_name: Pascanu, Razvan
last_name: Pascanu
- first_name: Adrian
full_name: Popescu, Adrian
last_name: Popescu
- first_name: Andreas S.
full_name: Tolias, Andreas S.
last_name: Tolias
- first_name: Joost
full_name: Van De Weijer, Joost
last_name: Van De Weijer
- first_name: Bing
full_name: Liu, Bing
last_name: Liu
- first_name: Vincenzo
full_name: Lomonaco, Vincenzo
last_name: Lomonaco
- first_name: Tinne
full_name: Tuytelaars, Tinne
last_name: Tuytelaars
- first_name: Gido M.
full_name: Van De Ven, Gido M.
last_name: Van De Ven
citation:
ama: 'Verwimp E, Aljundi R, Ben-David S, et al. Continual learning: Applications
and the road forward. Transactions on Machine Learning Research. 2024;2024.'
apa: 'Verwimp, E., Aljundi, R., Ben-David, S., Bethge, M., Cossu, A., Gepperth,
A., … Van De Ven, G. M. (2024). Continual learning: Applications and the road
forward. Transactions on Machine Learning Research. Transactions on Machine
Learning Research.'
chicago: 'Verwimp, Eli, Rahaf Aljundi, Shai Ben-David, Matthias Bethge, Andrea Cossu,
Alexander Gepperth, Tyler L. Hayes, et al. “Continual Learning: Applications and
the Road Forward.” Transactions on Machine Learning Research. Transactions
on Machine Learning Research, 2024.'
ieee: 'E. Verwimp et al., “Continual learning: Applications and the road
forward,” Transactions on Machine Learning Research, vol. 2024. Transactions
on Machine Learning Research, 2024.'
ista: 'Verwimp E, Aljundi R, Ben-David S, Bethge M, Cossu A, Gepperth A, Hayes TL,
Hüllermeier E, Kanan C, Kudithipudi D, Lampert C, Mundt M, Pascanu R, Popescu
A, Tolias AS, Van De Weijer J, Liu B, Lomonaco V, Tuytelaars T, Van De Ven GM.
2024. Continual learning: Applications and the road forward. Transactions on Machine
Learning Research. 2024.'
mla: 'Verwimp, Eli, et al. “Continual Learning: Applications and the Road Forward.”
Transactions on Machine Learning Research, vol. 2024, Transactions on Machine
Learning Research, 2024.'
short: E. Verwimp, R. Aljundi, S. Ben-David, M. Bethge, A. Cossu, A. Gepperth, T.L.
Hayes, E. Hüllermeier, C. Kanan, D. Kudithipudi, C. Lampert, M. Mundt, R. Pascanu,
A. Popescu, A.S. Tolias, J. Van De Weijer, B. Liu, V. Lomonaco, T. Tuytelaars,
G.M. Van De Ven, Transactions on Machine Learning Research 2024 (2024).
date_created: 2025-03-16T23:01:25Z
date_published: 2024-04-12T00:00:00Z
date_updated: 2025-03-20T09:21:02Z
day: '12'
ddc:
- '000'
department:
- _id: ChLa
external_id:
arxiv:
- '2311.11908'
file:
- access_level: open_access
checksum: 0714e12f7423cd098976ed9974561155
content_type: application/pdf
creator: dernst
date_created: 2025-03-20T09:02:18Z
date_updated: 2025-03-20T09:02:18Z
file_id: '19426'
file_name: 2024_TMLR_Verwimp.pdf
file_size: 1367966
relation: main_file
success: 1
file_date_updated: 2025-03-20T09:02:18Z
has_accepted_license: '1'
intvolume: ' 2024'
language:
- iso: eng
month: '04'
oa: 1
oa_version: Published Version
publication: Transactions on Machine Learning Research
publication_identifier:
eissn:
- 2835-8856
publication_status: published
publisher: Transactions on Machine Learning Research
quality_controlled: '1'
scopus_import: '1'
status: public
title: 'Continual learning: Applications and the road forward'
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: journal_article
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 2024
year: '2024'
...
---
_id: '17093'
abstract:
- lang: eng
text: 'Federated Learning (FL) enables large-scale distributed training of machine
learning models, while still allowing individual nodes to maintain data locally.
However, executing FL at scale comes with inherent practical challenges: 1) heterogeneity
of the local node data distributions, 2) heterogeneity of node computational speeds
(asynchrony), but also 3) constraints in the amount of communication between the
clients and the server. In this work, we present the first variant of the classic
federated averaging (FedAvg) algorithm which, at the same time, supports data
heterogeneity, partial client asynchrony, and communication compression. Our algorithm
comes with a novel, rigorous analysis showing that, in spite of these system relaxations,
it can provide similar convergence to FedAvg in interesting parameter regimes.
Experimental results in the rigorous LEAF benchmark on setups of up to 300 nodes
show that our algorithm ensures fast convergence for standard federated tasks,
improving upon prior quantized and asynchronous approaches.'
alternative_title:
- PMLR
article_processing_charge: No
arxiv: 1
author:
- first_name: Hossein
full_name: Zakerinia, Hossein
id: 653bd8b6-f394-11eb-9cf6-c0bbf6cd78d4
last_name: Zakerinia
- first_name: Shayan
full_name: Talaei, Shayan
last_name: Talaei
- first_name: Giorgi
full_name: Nadiradze, Giorgi
id: 3279A00C-F248-11E8-B48F-1D18A9856A87
last_name: Nadiradze
orcid: 0000-0001-5634-0731
- first_name: Dan-Adrian
full_name: Alistarh, Dan-Adrian
id: 4A899BFC-F248-11E8-B48F-1D18A9856A87
last_name: Alistarh
orcid: 0000-0003-3650-940X
citation:
ama: 'Zakerinia H, Talaei S, Nadiradze G, Alistarh D-A. Communication-efficient
federated learning with data and client heterogeneity. In: Proceedings of the
27th International Conference on Artificial Intelligence and Statistics. Vol
238. ML Research Press; 2024:3448-3456.'
apa: 'Zakerinia, H., Talaei, S., Nadiradze, G., & Alistarh, D.-A. (2024). Communication-efficient
federated learning with data and client heterogeneity. In Proceedings of the
27th International Conference on Artificial Intelligence and Statistics (Vol.
238, pp. 3448–3456). Valencia, Spain: ML Research Press.'
chicago: Zakerinia, Hossein, Shayan Talaei, Giorgi Nadiradze, and Dan-Adrian Alistarh.
“Communication-Efficient Federated Learning with Data and Client Heterogeneity.”
In Proceedings of the 27th International Conference on Artificial Intelligence
and Statistics, 238:3448–56. ML Research Press, 2024.
ieee: H. Zakerinia, S. Talaei, G. Nadiradze, and D.-A. Alistarh, “Communication-efficient
federated learning with data and client heterogeneity,” in Proceedings of the
27th International Conference on Artificial Intelligence and Statistics, Valencia,
Spain, 2024, vol. 238, pp. 3448–3456.
ista: 'Zakerinia H, Talaei S, Nadiradze G, Alistarh D-A. 2024. Communication-efficient
federated learning with data and client heterogeneity. Proceedings of the 27th
International Conference on Artificial Intelligence and Statistics. AISTATS: Conference
on Artificial Intelligence and Statistics, PMLR, vol. 238, 3448–3456.'
mla: Zakerinia, Hossein, et al. “Communication-Efficient Federated Learning with
Data and Client Heterogeneity.” Proceedings of the 27th International Conference
on Artificial Intelligence and Statistics, vol. 238, ML Research Press, 2024,
pp. 3448–56.
short: H. Zakerinia, S. Talaei, G. Nadiradze, D.-A. Alistarh, in:, Proceedings of
the 27th International Conference on Artificial Intelligence and Statistics, ML
Research Press, 2024, pp. 3448–3456.
conference:
end_date: 2024-05-04
location: Valencia, Spain
name: 'AISTATS: Conference on Artificial Intelligence and Statistics'
start_date: 2024-05-02
corr_author: '1'
date_created: 2024-06-02T22:00:57Z
date_published: 2024-05-01T00:00:00Z
date_updated: 2024-10-09T21:08:57Z
day: '01'
department:
- _id: DaAl
- _id: ChLa
external_id:
arxiv:
- '2206.10032'
intvolume: ' 238'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2206.10032
month: '05'
oa: 1
oa_version: Preprint
page: 3448-3456
publication: Proceedings of the 27th International Conference on Artificial Intelligence
and Statistics
publication_identifier:
eissn:
- 2640-3498
publication_status: published
publisher: ML Research Press
quality_controlled: '1'
scopus_import: '1'
status: public
title: Communication-efficient federated learning with data and client heterogeneity
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 238
year: '2024'
...
---
_id: '17411'
abstract:
- lang: eng
text: "We present PeFLL, a new personalized federated learning algorithm that improves\r\nover
the state-of-the-art in three aspects: 1) it produces more accurate models,\r\nespecially
in the low-data regime, and not only for clients present during its\r\ntraining
phase, but also for any that may emerge in the future; 2) it reduces the\r\namount
of on-client computation and client-server communication by providing\r\nfuture
clients with ready-to-use personalized models that require no additional\r\nfinetuning
or optimization; 3) it comes with theoretical guarantees that establish\r\ngeneralization
from the observed clients to future ones.\r\nAt the core of PeFLL lies a learning-to-learn
approach that jointly trains an\r\nembedding network and a hypernetwork. The embedding
network is used to\r\nrepresent clients in a latent descriptor space in a way
that reflects their similarity\r\nto each other. The hypernetwork takes as input
such descriptors and outputs the\r\nparameters of fully personalized client models.
In combination, both networks\r\nconstitute a learning algorithm that achieves
state-of-the-art performance in several\r\npersonalized federated learning benchmarks"
acknowledged_ssus:
- _id: ScienComp
acknowledgement: "This research was supported by the Scientific Service Units (SSU)
of ISTA through resources provided by Scientific Computing (SciComp).\r\n"
article_processing_charge: No
arxiv: 1
author:
- first_name: Jonathan A
full_name: Scott, Jonathan A
id: e499926b-f6e0-11ea-865d-9c63db0031e8
last_name: Scott
- first_name: Hossein
full_name: Zakerinia, Hossein
id: 653bd8b6-f394-11eb-9cf6-c0bbf6cd78d4
last_name: Zakerinia
orcid: 0009-0007-3977-6462
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Scott JA, Zakerinia H, Lampert C. PEFLL: Personalized federated learning by
learning to learn. In: 12th International Conference on Learning Representations.
OpenReview; 2024.'
apa: 'Scott, J. A., Zakerinia, H., & Lampert, C. (2024). PEFLL: Personalized
federated learning by learning to learn. In 12th International Conference on
Learning Representations. Vienna, Austria: OpenReview.'
chicago: 'Scott, Jonathan A, Hossein Zakerinia, and Christoph Lampert. “PEFLL: Personalized
Federated Learning by Learning to Learn.” In 12th International Conference
on Learning Representations. OpenReview, 2024.'
ieee: 'J. A. Scott, H. Zakerinia, and C. Lampert, “PEFLL: Personalized federated
learning by learning to learn,” in 12th International Conference on Learning
Representations, Vienna, Austria, 2024.'
ista: 'Scott JA, Zakerinia H, Lampert C. 2024. PEFLL: Personalized federated learning
by learning to learn. 12th International Conference on Learning Representations.
ICLR: International Conference on Learning Representations.'
mla: 'Scott, Jonathan A., et al. “PEFLL: Personalized Federated Learning by Learning
to Learn.” 12th International Conference on Learning Representations, OpenReview,
2024.'
short: J.A. Scott, H. Zakerinia, C. Lampert, in:, 12th International Conference
on Learning Representations, OpenReview, 2024.
conference:
end_date: 2024-03-07
location: Vienna, Austria
name: 'ICLR: International Conference on Learning Representations'
start_date: 2024-03-07
corr_author: '1'
date_created: 2024-08-11T22:01:12Z
date_published: 2024-03-07T00:00:00Z
date_updated: 2026-04-07T11:46:11Z
day: '07'
ddc:
- '000'
department:
- _id: ChLa
external_id:
arxiv:
- '2306.05515'
file:
- access_level: open_access
checksum: 81b7ea2e667adaf9c7a7b6b376b1f251
content_type: application/pdf
creator: dernst
date_created: 2024-08-12T07:38:06Z
date_updated: 2024-08-12T07:38:06Z
file_id: '17415'
file_name: 2024_ICLR_Scott.pdf
file_size: 1029219
relation: main_file
success: 1
file_date_updated: 2024-08-12T07:38:06Z
has_accepted_license: '1'
language:
- iso: eng
month: '03'
oa: 1
oa_version: Published Version
publication: 12th International Conference on Learning Representations
publication_status: published
publisher: OpenReview
quality_controlled: '1'
related_material:
record:
- id: '21198'
relation: dissertation_contains
status: public
scopus_import: '1'
status: public
title: 'PEFLL: Personalized federated learning by learning to learn'
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2024'
...
---
_id: '18120'
abstract:
- lang: eng
text: In practice, training using federated learning can be orders of magnitude
slower than standard centralized training. This severely limits the amount of
experimentation and tuning that can be done, making it challenging to obtain good
performance on a given task. Server-side proxy data can be used to run training
simulations, for instance for hyperparameter tuning. This can greatly speed up
the training pipeline by reducing the number of tuning runs to be performed overall
on the true clients. However, it is challenging to ensure that these simulations
accurately reflect the dynamics of the real federated training. In particular,
the proxy data used for simulations often comes as a single centralized dataset
without a partition into distinct clients, and partitioning this data in a naive
way can lead to simulations that poorly reflect real federated training. In this
paper we address the challenge of how to partition centralized data in a way that
reflects the statistical heterogeneity of the true federated clients. We propose
a fully federated, theoretically justified, algorithm that efficiently learns
the distribution of the true clients and observe improved server-side simulations
when using the inferred distribution to create simulated clients from the centralized
data.
acknowledgement: 'We would like to thank: Mona Chitnis and everyone in the Private
Federated Learning team at Apple for their help and support throughout the entire
project; Audra McMillan, Martin Pelikan, Anosh Raj and Barry Theobold for feedback
on the initial versions of the paper; and Christoph Lampert for valuable feedback
on the paper structure and suggestions for additional experiments.'
alternative_title:
- PMLR
article_processing_charge: No
arxiv: 1
author:
- first_name: Jonathan A
full_name: Scott, Jonathan A
id: e499926b-f6e0-11ea-865d-9c63db0031e8
last_name: Scott
- first_name: Áine
full_name: Cahill, Áine
last_name: Cahill
citation:
ama: 'Scott JA, Cahill Á. Improved modelling of federated datasets using mixtures-of-Dirichlet-multinomials.
In: Proceedings of the 41st International Conference on Machine Learning.
Vol 235. ML Research Press; 2024:44012-44037.'
apa: 'Scott, J. A., & Cahill, Á. (2024). Improved modelling of federated datasets
using mixtures-of-Dirichlet-multinomials. In Proceedings of the 41st International
Conference on Machine Learning (Vol. 235, pp. 44012–44037). Vienna, Austria:
ML Research Press.'
chicago: Scott, Jonathan A, and Áine Cahill. “Improved Modelling of Federated Datasets
Using Mixtures-of-Dirichlet-Multinomials.” In Proceedings of the 41st International
Conference on Machine Learning, 235:44012–37. ML Research Press, 2024.
ieee: J. A. Scott and Á. Cahill, “Improved modelling of federated datasets using
mixtures-of-Dirichlet-multinomials,” in Proceedings of the 41st International
Conference on Machine Learning, Vienna, Austria, 2024, vol. 235, pp. 44012–44037.
ista: 'Scott JA, Cahill Á. 2024. Improved modelling of federated datasets using
mixtures-of-Dirichlet-multinomials. Proceedings of the 41st International Conference
on Machine Learning. ICML: International Conference on Machine Learning, PMLR,
vol. 235, 44012–44037.'
mla: Scott, Jonathan A., and Áine Cahill. “Improved Modelling of Federated Datasets
Using Mixtures-of-Dirichlet-Multinomials.” Proceedings of the 41st International
Conference on Machine Learning, vol. 235, ML Research Press, 2024, pp. 44012–37.
short: J.A. Scott, Á. Cahill, in:, Proceedings of the 41st International Conference
on Machine Learning, ML Research Press, 2024, pp. 44012–44037.
conference:
end_date: 2024-07-27
location: Vienna, Austria
name: 'ICML: International Conference on Machine Learning'
start_date: 2024-07-21
corr_author: '1'
date_created: 2024-09-22T22:01:45Z
date_published: 2024-09-01T00:00:00Z
date_updated: 2026-04-07T11:46:11Z
day: '01'
department:
- _id: ChLa
external_id:
arxiv:
- '2406.02416'
intvolume: ' 235'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2406.02416
month: '09'
oa: 1
oa_version: Preprint
page: 44012-44037
publication: Proceedings of the 41st International Conference on Machine Learning
publication_identifier:
eissn:
- 2640-3498
publication_status: published
publisher: ML Research Press
quality_controlled: '1'
related_material:
record:
- id: '21198'
relation: dissertation_contains
status: public
scopus_import: '1'
status: public
title: Improved modelling of federated datasets using mixtures-of-Dirichlet-multinomials
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 235
year: '2024'
...
---
OA_place: repository
OA_type: green
_id: '17426'
abstract:
- lang: eng
text: "The robustness of neural networks against input perturbations with bounded\r\nmagnitude
represents a serious concern in the deployment of deep learning\r\nmodels in safety-critical
systems. Recently, the scientific community has\r\nfocused on enhancing certifiable
robustness guarantees by crafting 1-Lipschitz\r\nneural networks that leverage
Lipschitz bounded dense and convolutional layers.\r\nAlthough different methods
have been proposed in the literature to achieve this\r\ngoal, understanding the
performance of such methods is not straightforward,\r\nsince different metrics
can be relevant (e.g., training time, memory usage,\r\naccuracy, certifiable robustness)
for different applications. For this reason,\r\nthis work provides a thorough
theoretical and empirical comparison between\r\nmethods by evaluating them in
terms of memory usage, speed, and certifiable\r\nrobust accuracy. The paper also
provides some guidelines and recommendations to\r\nsupport the user in selecting
the methods that work best depending on the\r\navailable resources. We provide
code at\r\nhttps://github.com/berndprach/1LipschitzLayersCompared."
acknowledgement: "This work was partially supported by project SERICS (PE00000014)
under the MUR National Recovery and Resilience Plan funded by the European Union
- NextGenerationEU.\r\n"
article_processing_charge: No
arxiv: 1
author:
- first_name: Bernd
full_name: Prach, Bernd
id: 2D561D42-C427-11E9-89B4-9C1AE6697425
last_name: Prach
- first_name: Fabio
full_name: Brau, Fabio
last_name: Brau
- first_name: Giorgio
full_name: Buttazzo, Giorgio
last_name: Buttazzo
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Prach B, Brau F, Buttazzo G, Lampert C. 1-Lipschitz layers compared: Memory,
speed, and certifiable robustness. In: Proceedings of the IEEE/CVF Conference
on Computer Vision and Pattern Recognition. Computer Vision Foundation; 2024:24574-24583.
doi:10.1109/CVPR52733.2024.02320'
apa: 'Prach, B., Brau, F., Buttazzo, G., & Lampert, C. (2024). 1-Lipschitz layers
compared: Memory, speed, and certifiable robustness. In Proceedings of the
IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 24574–24583).
Seattle, WA, United States: Computer Vision Foundation. https://doi.org/10.1109/CVPR52733.2024.02320'
chicago: 'Prach, Bernd, Fabio Brau, Giorgio Buttazzo, and Christoph Lampert. “1-Lipschitz
Layers Compared: Memory, Speed, and Certifiable Robustness.” In Proceedings
of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 24574–83.
Computer Vision Foundation, 2024. https://doi.org/10.1109/CVPR52733.2024.02320.'
ieee: 'B. Prach, F. Brau, G. Buttazzo, and C. Lampert, “1-Lipschitz layers compared:
Memory, speed, and certifiable robustness,” in Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition, Seattle, WA, United
States, 2024, pp. 24574–24583.'
ista: 'Prach B, Brau F, Buttazzo G, Lampert C. 2024. 1-Lipschitz layers compared:
Memory, speed, and certifiable robustness. Proceedings of the IEEE/CVF Conference
on Computer Vision and Pattern Recognition. CVPR: Conference on Computer Vision
and Pattern Recognition, 24574–24583.'
mla: 'Prach, Bernd, et al. “1-Lipschitz Layers Compared: Memory, Speed, and Certifiable
Robustness.” Proceedings of the IEEE/CVF Conference on Computer Vision and
Pattern Recognition, Computer Vision Foundation, 2024, pp. 24574–83, doi:10.1109/CVPR52733.2024.02320.'
short: B. Prach, F. Brau, G. Buttazzo, C. Lampert, in:, Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition, Computer Vision Foundation,
2024, pp. 24574–24583.
conference:
end_date: 2024-06-22
location: Seattle, WA, United States
name: 'CVPR: Conference on Computer Vision and Pattern Recognition'
start_date: 2024-06-16
corr_author: '1'
date_created: 2024-08-14T08:42:32Z
date_published: 2024-06-01T00:00:00Z
date_updated: 2026-04-07T11:49:51Z
day: '01'
department:
- _id: GradSch
- _id: ChLa
doi: 10.1109/CVPR52733.2024.02320
external_id:
arxiv:
- '2311.16833'
isi:
- '001344387500055'
has_accepted_license: '1'
isi: 1
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2311.16833
month: '06'
oa: 1
oa_version: Preprint
page: 24574-24583
publication: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern
Recognition
publication_status: published
publisher: Computer Vision Foundation
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://github.com/berndprach/1LipschitzLayersCompared
record:
- id: '19759'
relation: dissertation_contains
status: public
status: public
title: '1-Lipschitz layers compared: Memory, speed, and certifiable robustness'
type: conference
user_id: 317138e5-6ab7-11ef-aa6d-ffef3953e345
year: '2024'
...
---
OA_place: repository
_id: '18874'
abstract:
- lang: eng
text: "Despite extensive research since the community learned about adversarial\r\nexamples
10 years ago, we still do not know how to train high-accuracy\r\nclassifiers that
are guaranteed to be robust to small perturbations of their\r\ninputs. Previous
works often argued that this might be because no classifier\r\nexists that is
robust and accurate at the same time. However, in computer\r\nvision this assumption
does not match reality where humans are usually accurate\r\nand robust on most
tasks of interest. We offer an alternative explanation and\r\nshow that in certain
settings robust generalization is only possible with\r\nunrealistically large
amounts of data. More precisely we find a setting where a\r\nrobust classifier
exists, it is easy to learn an accurate classifier, yet it\r\nrequires an exponential
amount of data to learn a robust classifier. Based on\r\nthis theoretical result,
we explore how well robust classifiers generalize on\r\ndatasets such as CIFAR-10.
We come to the conclusion that on this datasets, the\r\nlimitation of current
robust models also lies in the generalization, and that\r\nthey require a lot
of data to do well on the test set. We also show that the\r\nproblem is not in
the expressiveness or generalization capabilities of current\r\narchitectures,
and that there are low magnitude features in the data which are\r\nuseful for
non-robust generalization but are not available for robust\r\nclassifiers."
article_number: '2412.04245'
article_processing_charge: No
arxiv: 1
author:
- first_name: Bernd
full_name: Prach, Bernd
id: 2D561D42-C427-11E9-89B4-9C1AE6697425
last_name: Prach
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: Prach B, Lampert C. Intriguing properties of robust classification. arXiv.
doi:10.48550/arXiv.2412.04245
apa: Prach, B., & Lampert, C. (n.d.). Intriguing properties of robust classification.
arXiv. https://doi.org/10.48550/arXiv.2412.04245
chicago: Prach, Bernd, and Christoph Lampert. “Intriguing Properties of Robust Classification.”
ArXiv, n.d. https://doi.org/10.48550/arXiv.2412.04245.
ieee: B. Prach and C. Lampert, “Intriguing properties of robust classification,”
arXiv. .
ista: Prach B, Lampert C. Intriguing properties of robust classification. arXiv,
2412.04245.
mla: Prach, Bernd, and Christoph Lampert. “Intriguing Properties of Robust Classification.”
ArXiv, 2412.04245, doi:10.48550/arXiv.2412.04245.
short: B. Prach, C. Lampert, ArXiv (n.d.).
corr_author: '1'
date_created: 2025-01-24T16:57:29Z
date_published: 2024-12-05T00:00:00Z
date_updated: 2026-04-07T11:49:51Z
day: '05'
department:
- _id: GradSch
- _id: ChLa
doi: 10.48550/arXiv.2412.04245
external_id:
arxiv:
- '2412.04245'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2412.04245
month: '12'
oa: 1
oa_version: Preprint
publication: arXiv
publication_status: draft
related_material:
record:
- id: '20455'
relation: later_version
status: public
- id: '19759'
relation: dissertation_contains
status: public
status: public
title: Intriguing properties of robust classification
type: preprint
user_id: 8b945eb4-e2f2-11eb-945a-df72226e66a9
year: '2024'
...
---
_id: '14320'
abstract:
- lang: eng
text: The development of two-dimensional materials has resulted in a diverse range
of novel, high-quality compounds with increasing complexity. A key requirement
for a comprehensive quantitative theory is the accurate determination of these
materials' band structure parameters. However, this task is challenging due to
the intricate band structures and the indirect nature of experimental probes.
In this work, we introduce a general framework to derive band structure parameters
from experimental data using deep neural networks. We applied our method to the
penetration field capacitance measurement of trilayer graphene, an effective probe
of its density of states. First, we demonstrate that a trained deep network gives
accurate predictions for the penetration field capacitance as a function of tight-binding
parameters. Next, we use the fast and accurate predictions from the trained network
to automatically determine tight-binding parameters directly from experimental
data, with extracted parameters being in a good agreement with values in the literature.
We conclude by discussing potential applications of our method to other materials
and experimental techniques beyond penetration field capacitance.
acknowledgement: A.F.Y. acknowledges primary support from the Department of Energy
under award DE-SC0020043, and additional support from the Gordon and Betty Moore
Foundation under award GBMF9471 for group operations.
article_number: '125411'
article_processing_charge: No
article_type: original
arxiv: 1
author:
- first_name: Paul M
full_name: Henderson, Paul M
id: 13C09E74-18D9-11E9-8878-32CFE5697425
last_name: Henderson
orcid: 0000-0002-5198-7445
- first_name: Areg
full_name: Ghazaryan, Areg
id: 4AF46FD6-F248-11E8-B48F-1D18A9856A87
last_name: Ghazaryan
orcid: 0000-0001-9666-3543
- first_name: Alexander A.
full_name: Zibrov, Alexander A.
last_name: Zibrov
- first_name: Andrea F.
full_name: Young, Andrea F.
last_name: Young
- first_name: Maksym
full_name: Serbyn, Maksym
id: 47809E7E-F248-11E8-B48F-1D18A9856A87
last_name: Serbyn
orcid: 0000-0002-2399-5827
citation:
ama: 'Henderson PM, Ghazaryan A, Zibrov AA, Young AF, Serbyn M. Deep learning extraction
of band structure parameters from density of states: A case study on trilayer
graphene. Physical Review B. 2023;108(12). doi:10.1103/physrevb.108.125411'
apa: 'Henderson, P. M., Ghazaryan, A., Zibrov, A. A., Young, A. F., & Serbyn,
M. (2023). Deep learning extraction of band structure parameters from density
of states: A case study on trilayer graphene. Physical Review B. American
Physical Society. https://doi.org/10.1103/physrevb.108.125411'
chicago: 'Henderson, Paul M, Areg Ghazaryan, Alexander A. Zibrov, Andrea F. Young,
and Maksym Serbyn. “Deep Learning Extraction of Band Structure Parameters from
Density of States: A Case Study on Trilayer Graphene.” Physical Review B.
American Physical Society, 2023. https://doi.org/10.1103/physrevb.108.125411.'
ieee: 'P. M. Henderson, A. Ghazaryan, A. A. Zibrov, A. F. Young, and M. Serbyn,
“Deep learning extraction of band structure parameters from density of states:
A case study on trilayer graphene,” Physical Review B, vol. 108, no. 12.
American Physical Society, 2023.'
ista: 'Henderson PM, Ghazaryan A, Zibrov AA, Young AF, Serbyn M. 2023. Deep learning
extraction of band structure parameters from density of states: A case study on
trilayer graphene. Physical Review B. 108(12), 125411.'
mla: 'Henderson, Paul M., et al. “Deep Learning Extraction of Band Structure Parameters
from Density of States: A Case Study on Trilayer Graphene.” Physical Review
B, vol. 108, no. 12, 125411, American Physical Society, 2023, doi:10.1103/physrevb.108.125411.'
short: P.M. Henderson, A. Ghazaryan, A.A. Zibrov, A.F. Young, M. Serbyn, Physical
Review B 108 (2023).
date_created: 2023-09-12T07:12:12Z
date_published: 2023-09-15T00:00:00Z
date_updated: 2023-09-20T09:38:24Z
day: '15'
department:
- _id: MaSe
- _id: ChLa
- _id: MiLe
doi: 10.1103/physrevb.108.125411
external_id:
arxiv:
- '2210.06310'
intvolume: ' 108'
issue: '12'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2210.06310
month: '09'
oa: 1
oa_version: Preprint
publication: Physical Review B
publication_identifier:
eissn:
- 2469-9969
issn:
- 2469-9950
publication_status: published
publisher: American Physical Society
quality_controlled: '1'
scopus_import: '1'
status: public
title: 'Deep learning extraction of band structure parameters from density of states:
A case study on trilayer graphene'
type: journal_article
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 108
year: '2023'
...
---
_id: '14410'
abstract:
- lang: eng
text: This paper focuses on the implementation details of the baseline methods and
a recent lightweight conditional model extrapolation algorithm LIMES [5] for streaming
data under class-prior shift. LIMES achieves superior performance over the baseline
methods, especially concerning the minimum-across-day accuracy, which is important
for the users of the system. In this work, the key measures to facilitate reproducibility
and enhance the credibility of the results are described.
alternative_title:
- LNCS
article_processing_charge: No
author:
- first_name: Paulina
full_name: Tomaszewska, Paulina
last_name: Tomaszewska
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Tomaszewska P, Lampert C. On the implementation of baselines and lightweight
conditional model extrapolation (LIMES) under class-prior shift. In: International
Workshop on Reproducible Research in Pattern Recognition. Vol 14068. Springer
Nature; 2023:67-73. doi:10.1007/978-3-031-40773-4_6'
apa: 'Tomaszewska, P., & Lampert, C. (2023). On the implementation of baselines
and lightweight conditional model extrapolation (LIMES) under class-prior shift.
In International Workshop on Reproducible Research in Pattern Recognition
(Vol. 14068, pp. 67–73). Montreal, Canada: Springer Nature. https://doi.org/10.1007/978-3-031-40773-4_6'
chicago: Tomaszewska, Paulina, and Christoph Lampert. “On the Implementation of Baselines
and Lightweight Conditional Model Extrapolation (LIMES) under Class-Prior Shift.”
In International Workshop on Reproducible Research in Pattern Recognition,
14068:67–73. Springer Nature, 2023. https://doi.org/10.1007/978-3-031-40773-4_6.
ieee: P. Tomaszewska and C. Lampert, “On the implementation of baselines and lightweight
conditional model extrapolation (LIMES) under class-prior shift,” in International
Workshop on Reproducible Research in Pattern Recognition, Montreal, Canada,
2023, vol. 14068, pp. 67–73.
ista: 'Tomaszewska P, Lampert C. 2023. On the implementation of baselines and lightweight
conditional model extrapolation (LIMES) under class-prior shift. International
Workshop on Reproducible Research in Pattern Recognition. RRPR: Reproducible Research
in Pattern Recognition, LNCS, vol. 14068, 67–73.'
mla: Tomaszewska, Paulina, and Christoph Lampert. “On the Implementation of Baselines
and Lightweight Conditional Model Extrapolation (LIMES) under Class-Prior Shift.”
International Workshop on Reproducible Research in Pattern Recognition,
vol. 14068, Springer Nature, 2023, pp. 67–73, doi:10.1007/978-3-031-40773-4_6.
short: P. Tomaszewska, C. Lampert, in:, International Workshop on Reproducible Research
in Pattern Recognition, Springer Nature, 2023, pp. 67–73.
conference:
end_date: 2022-08-21
location: Montreal, Canada
name: 'RRPR: Reproducible Research in Pattern Recognition'
start_date: 2022-08-21
date_created: 2023-10-08T22:01:18Z
date_published: 2023-08-20T00:00:00Z
date_updated: 2023-10-09T06:48:02Z
day: '20'
department:
- _id: ChLa
doi: 10.1007/978-3-031-40773-4_6
intvolume: ' 14068'
language:
- iso: eng
month: '08'
oa_version: None
page: 67-73
publication: International Workshop on Reproducible Research in Pattern Recognition
publication_identifier:
eissn:
- 1611-3349
isbn:
- '9783031407727'
issn:
- 0302-9743
publication_status: published
publisher: Springer Nature
quality_controlled: '1'
scopus_import: '1'
status: public
title: On the implementation of baselines and lightweight conditional model extrapolation
(LIMES) under class-prior shift
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 14068
year: '2023'
...
---
_id: '14446'
abstract:
- lang: eng
text: Recent work has paid close attention to the first principle of Granger causality,
according to which cause precedes effect. In this context, the question may arise
whether the detected direction of causality also reverses after the time reversal
of unidirectionally coupled data. Recently, it has been shown that for unidirectionally
causally connected autoregressive (AR) processes X → Y, after time reversal of
data, the opposite causal direction Y → X is indeed detected, although typically
as part of the bidirectional X↔ Y link. As we argue here, the answer is different
when the measured data are not from AR processes but from linked deterministic
systems. When the goal is the usual forward data analysis, cross-mapping-like
approaches correctly detect X → Y, while Granger causality-like approaches, which
should not be used for deterministic time series, detect causal independence X
→ Y. The results of backward causal analysis depend on the predictability of the
reversed data. Unlike AR processes, observables from deterministic dynamical systems,
even complex nonlinear ones, can be predicted well forward, while backward predictions
can be difficult (notably when the time reversal of a function leads to one-to-many
relations). To address this problem, we propose an approach based on models that
provide multiple candidate predictions for the target, combined with a loss function
that consideres only the best candidate. The resulting good forward and backward
predictability supports the view that unidirectionally causally linked deterministic
dynamical systems X → Y can be expected to detect the same link both before and
after time reversal.
acknowledgement: The work was supported by the Scientific Grant Agency of the Ministry
of Education of the Slovak Republic and the Slovak Academy of Sciences, projects
APVV-21-0216, VEGA2-0096-21 and VEGA 2-0023-22.
article_processing_charge: Yes
article_type: original
author:
- first_name: Jozef
full_name: Jakubík, Jozef
last_name: Jakubík
- first_name: Phuong
full_name: Bui Thi Mai, Phuong
id: 3EC6EE64-F248-11E8-B48F-1D18A9856A87
last_name: Bui Thi Mai
- first_name: Martina
full_name: Chvosteková, Martina
last_name: Chvosteková
- first_name: Anna
full_name: Krakovská, Anna
last_name: Krakovská
citation:
ama: Jakubík J, Phuong M, Chvosteková M, Krakovská A. Against the flow of time with
multi-output models. Measurement Science Review. 2023;23(4):175-183. doi:10.2478/msr-2023-0023
apa: Jakubík, J., Phuong, M., Chvosteková, M., & Krakovská, A. (2023). Against
the flow of time with multi-output models. Measurement Science Review.
Sciendo. https://doi.org/10.2478/msr-2023-0023
chicago: Jakubík, Jozef, Mary Phuong, Martina Chvosteková, and Anna Krakovská. “Against
the Flow of Time with Multi-Output Models.” Measurement Science Review.
Sciendo, 2023. https://doi.org/10.2478/msr-2023-0023.
ieee: J. Jakubík, M. Phuong, M. Chvosteková, and A. Krakovská, “Against the flow
of time with multi-output models,” Measurement Science Review, vol. 23,
no. 4. Sciendo, pp. 175–183, 2023.
ista: Jakubík J, Phuong M, Chvosteková M, Krakovská A. 2023. Against the flow of
time with multi-output models. Measurement Science Review. 23(4), 175–183.
mla: Jakubík, Jozef, et al. “Against the Flow of Time with Multi-Output Models.”
Measurement Science Review, vol. 23, no. 4, Sciendo, 2023, pp. 175–83,
doi:10.2478/msr-2023-0023.
short: J. Jakubík, M. Phuong, M. Chvosteková, A. Krakovská, Measurement Science
Review 23 (2023) 175–183.
date_created: 2023-10-22T22:01:15Z
date_published: 2023-08-01T00:00:00Z
date_updated: 2025-09-09T13:10:30Z
day: '01'
ddc:
- '510'
department:
- _id: ChLa
doi: 10.2478/msr-2023-0023
external_id:
isi:
- '001070829600005'
file:
- access_level: open_access
checksum: b069cc10fa6a7c96b2bc9f728165f9e6
content_type: application/pdf
creator: dernst
date_created: 2023-10-31T12:07:23Z
date_updated: 2023-10-31T12:07:23Z
file_id: '14476'
file_name: 2023_MeasurementScienceRev_Jakubik.pdf
file_size: 2639783
relation: main_file
success: 1
file_date_updated: 2023-10-31T12:07:23Z
has_accepted_license: '1'
intvolume: ' 23'
isi: 1
issue: '4'
language:
- iso: eng
license: https://creativecommons.org/licenses/by-nc-nd/4.0/
month: '08'
oa: 1
oa_version: Published Version
page: 175-183
publication: Measurement Science Review
publication_identifier:
eissn:
- 1335-8871
publication_status: published
publisher: Sciendo
quality_controlled: '1'
scopus_import: '1'
status: public
title: Against the flow of time with multi-output models
tmp:
image: /images/cc_by_nc_nd.png
legal_code_url: https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode
name: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
(CC BY-NC-ND 4.0)
short: CC BY-NC-ND (4.0)
type: journal_article
user_id: 317138e5-6ab7-11ef-aa6d-ffef3953e345
volume: 23
year: '2023'
...
---
OA_place: repository
OA_type: green
_id: '12660'
abstract:
- lang: eng
text: 'We present Cross-Client Label Propagation(XCLP), a new method for transductive
federated learning. XCLP estimates a data graph jointly from the data of multiple
clients and computes labels for the unlabeled data by propagating label information
across the graph. To avoid clients having to share their data with anyone, XCLP
employs two cryptographically secure protocols: secure Hamming distance computation
and secure summation. We demonstrate two distinct applications of XCLP within
federated learning. In the first, we use it in a one-shot way to predict labels
for unseen test points. In the second, we use it to repeatedly pseudo-label unlabeled
training data in a federated semi-supervised setting. Experiments on both real
federated and standard benchmark datasets show that in both applications XCLP
achieves higher classification accuracy than alternative approaches.'
alternative_title:
- TMLR
article_processing_charge: No
arxiv: 1
author:
- first_name: Jonathan A
full_name: Scott, Jonathan A
id: e499926b-f6e0-11ea-865d-9c63db0031e8
last_name: Scott
- first_name: Michelle X
full_name: Yeo, Michelle X
id: 2D82B818-F248-11E8-B48F-1D18A9856A87
last_name: Yeo
orcid: 0009-0001-3676-4809
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Scott JA, Yeo MX, Lampert C. Cross-client label propagation for transductive
and semi-supervised federated learning. In: Transactions in Machine Learning.
Curran Associates; 2023.'
apa: Scott, J. A., Yeo, M. X., & Lampert, C. (2023). Cross-client label propagation
for transductive and semi-supervised federated learning. In Transactions in
Machine Learning. Curran Associates.
chicago: Scott, Jonathan A, Michelle X Yeo, and Christoph Lampert. “Cross-Client
Label Propagation for Transductive and Semi-Supervised Federated Learning.” In
Transactions in Machine Learning. Curran Associates, 2023.
ieee: J. A. Scott, M. X. Yeo, and C. Lampert, “Cross-client label propagation for
transductive and semi-supervised federated learning,” in Transactions in Machine
Learning, 2023.
ista: Scott JA, Yeo MX, Lampert C. 2023. Cross-client label propagation for transductive
and semi-supervised federated learning. Transactions in Machine Learning. , TMLR,
.
mla: Scott, Jonathan A., et al. “Cross-Client Label Propagation for Transductive
and Semi-Supervised Federated Learning.” Transactions in Machine Learning,
Curran Associates, 2023.
short: J.A. Scott, M.X. Yeo, C. Lampert, in:, Transactions in Machine Learning,
Curran Associates, 2023.
corr_author: '1'
date_created: 2023-02-20T08:21:50Z
date_published: 2023-11-27T00:00:00Z
date_updated: 2025-02-04T08:32:19Z
day: '27'
ddc:
- '004'
department:
- _id: ChLa
external_id:
arxiv:
- '2210.06434'
file:
- access_level: open_access
checksum: aa322ad91cbd229f5cafe6733a119bd1
content_type: application/pdf
creator: dernst
date_created: 2025-02-04T08:30:05Z
date_updated: 2025-02-04T08:30:05Z
file_id: '18990'
file_name: 2023_TMLR_Scott.pdf
file_size: 553717
relation: main_file
success: 1
file_date_updated: 2025-02-04T08:30:05Z
has_accepted_license: '1'
language:
- iso: eng
month: '11'
oa: 1
oa_version: Preprint
publication: Transactions in Machine Learning
publication_identifier:
issn:
- 2835-8856
publication_status: published
publisher: Curran Associates
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://github.com/jonnyascott/xclp
status: public
title: Cross-client label propagation for transductive and semi-supervised federated
learning
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2023'
...
---
_id: '14921'
abstract:
- lang: eng
text: Neural collapse (NC) refers to the surprising structure of the last layer
of deep neural networks in the terminal phase of gradient descent training. Recently,
an increasing amount of experimental evidence has pointed to the propagation of
NC to earlier layers of neural networks. However, while the NC in the last layer
is well studied theoretically, much less is known about its multi-layered counterpart
- deep neural collapse (DNC). In particular, existing work focuses either on linear
layers or only on the last two layers at the price of an extra assumption. Our
paper fills this gap by generalizing the established analytical framework for
NC - the unconstrained features model - to multiple non-linear layers. Our key
technical contribution is to show that, in a deep unconstrained features model,
the unique global optimum for binary classification exhibits all the properties
typical of DNC. This explains the existing experimental evidence of DNC. We also
empirically show that (i) by optimizing deep unconstrained features models via
gradient descent, the resulting solution agrees well with our theory, and (ii)
trained networks recover the unconstrained features suitable for the occurrence
of DNC, thus supporting the validity of this modeling principle.
acknowledgement: M. M. is partially supported by the 2019 Lopez-Loreta Prize. The
authors would like to thank Eugenia Iofinova, Bernd Prach and Simone Bombari for
valuable feedback on the manuscript.
alternative_title:
- NeurIPS
article_processing_charge: No
arxiv: 1
author:
- first_name: Peter
full_name: Súkeník, Peter
id: d64d6a8d-eb8e-11eb-b029-96fd216dec3c
last_name: Súkeník
- first_name: Marco
full_name: Mondelli, Marco
id: 27EB676C-8706-11E9-9510-7717E6697425
last_name: Mondelli
orcid: 0000-0002-3242-7020
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: 'Súkeník P, Mondelli M, Lampert C. Deep neural collapse is provably optimal
for the deep unconstrained features model. In: 37th Annual Conference on Neural
Information Processing Systems. ; 2023.'
apa: Súkeník, P., Mondelli, M., & Lampert, C. (2023). Deep neural collapse is
provably optimal for the deep unconstrained features model. In 37th Annual
Conference on Neural Information Processing Systems. New Orleans, LA, United
States.
chicago: Súkeník, Peter, Marco Mondelli, and Christoph Lampert. “Deep Neural Collapse
Is Provably Optimal for the Deep Unconstrained Features Model.” In 37th Annual
Conference on Neural Information Processing Systems, 2023.
ieee: P. Súkeník, M. Mondelli, and C. Lampert, “Deep neural collapse is provably
optimal for the deep unconstrained features model,” in 37th Annual Conference
on Neural Information Processing Systems, New Orleans, LA, United States,
2023.
ista: 'Súkeník P, Mondelli M, Lampert C. 2023. Deep neural collapse is provably
optimal for the deep unconstrained features model. 37th Annual Conference on Neural
Information Processing Systems. NeurIPS: Neural Information Processing Systems,
NeurIPS, .'
mla: Súkeník, Peter, et al. “Deep Neural Collapse Is Provably Optimal for the Deep
Unconstrained Features Model.” 37th Annual Conference on Neural Information
Processing Systems, 2023.
short: P. Súkeník, M. Mondelli, C. Lampert, in:, 37th Annual Conference on Neural
Information Processing Systems, 2023.
conference:
end_date: 2023-12-16
location: New Orleans, LA, United States
name: 'NeurIPS: Neural Information Processing Systems'
start_date: 2023-12-10
corr_author: '1'
date_created: 2024-02-02T11:17:41Z
date_published: 2023-12-15T00:00:00Z
date_updated: 2025-04-15T07:50:16Z
day: '15'
department:
- _id: MaMo
- _id: ChLa
external_id:
arxiv:
- '2305.13165'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: ' https://doi.org/10.48550/arXiv.2305.13165'
month: '12'
oa: 1
oa_version: Preprint
project:
- _id: 059876FA-7A3F-11EA-A408-12923DDC885E
name: Prix Lopez-Loretta 2019 - Marco Mondelli
publication: 37th Annual Conference on Neural Information Processing Systems
publication_status: published
quality_controlled: '1'
status: public
title: Deep neural collapse is provably optimal for the deep unconstrained features
model
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2023'
...
---
OA_place: repository
_id: '15039'
abstract:
- lang: eng
text: 'A crucial property for achieving secure, trustworthy and interpretable deep
learning systems is their robustness: small changes to a system''s inputs should
not result in large changes to its outputs. Mathematically, this means one strives
for networks with a small Lipschitz constant. Several recent works have focused
on how to construct such Lipschitz networks, typically by imposing constraints
on the weight matrices. In this work, we study an orthogonal aspect, namely the
role of the activation function. We show that commonly used activation functions,
such as MaxMin, as well as all piece-wise linear ones with two segments unnecessarily
restrict the class of representable functions, even in the simplest one-dimensional
setting. We furthermore introduce the new N-activation function that is provably
more expressive than currently popular activation functions. We provide code at
this https URL.'
article_number: '2311.06103'
article_processing_charge: No
arxiv: 1
author:
- first_name: Bernd
full_name: Prach, Bernd
id: 2D561D42-C427-11E9-89B4-9C1AE6697425
last_name: Prach
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
citation:
ama: Prach B, Lampert C. 1-Lipschitz neural networks are more expressive with N-activations.
arXiv. doi:10.48550/ARXIV.2311.06103
apa: Prach, B., & Lampert, C. (n.d.). 1-Lipschitz neural networks are more expressive
with N-activations. arXiv. https://doi.org/10.48550/ARXIV.2311.06103
chicago: Prach, Bernd, and Christoph Lampert. “1-Lipschitz Neural Networks Are More
Expressive with N-Activations.” ArXiv, n.d. https://doi.org/10.48550/ARXIV.2311.06103.
ieee: B. Prach and C. Lampert, “1-Lipschitz neural networks are more expressive
with N-activations,” arXiv. .
ista: Prach B, Lampert C. 1-Lipschitz neural networks are more expressive with N-activations.
arXiv, 2311.06103.
mla: Prach, Bernd, and Christoph Lampert. “1-Lipschitz Neural Networks Are More
Expressive with N-Activations.” ArXiv, 2311.06103, doi:10.48550/ARXIV.2311.06103.
short: B. Prach, C. Lampert, ArXiv (n.d.).
corr_author: '1'
date_created: 2024-02-28T17:59:32Z
date_published: 2023-11-10T00:00:00Z
date_updated: 2026-04-07T11:49:51Z
day: '10'
department:
- _id: GradSch
- _id: ChLa
doi: 10.48550/ARXIV.2311.06103
external_id:
arxiv:
- '2311.06103'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2311.06103
month: '11'
oa: 1
oa_version: Preprint
publication: arXiv
publication_status: draft
related_material:
record:
- id: '19759'
relation: dissertation_contains
status: public
status: public
title: 1-Lipschitz neural networks are more expressive with N-activations
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: preprint
user_id: 8b945eb4-e2f2-11eb-945a-df72226e66a9
year: '2023'
...
---
OA_place: publisher
_id: '13074'
abstract:
- lang: eng
text: "Deep learning has become an integral part of a large number of important
applications, and many of the recent breakthroughs have been enabled by the ability
to train very large models, capable to capture complex patterns and relationships
from the data. At the same time, the massive sizes of modern deep learning models
have made their deployment to smaller devices more challenging; this is particularly
important, as in many applications the users rely on accurate deep learning predictions,
but they only have access to devices with limited memory and compute power. One
solution to this problem is to prune neural networks, by setting as many of their
parameters as possible to zero, to obtain accurate sparse models with lower memory
footprint. Despite the great research progress in obtaining sparse models that
preserve accuracy, while satisfying memory and computational constraints, there
are still many challenges associated with efficiently training sparse models,
as well as understanding their generalization properties.\r\n\r\nThe focus of
this thesis is to investigate how the training process of sparse models can be
made more efficient, and to understand the differences between sparse and dense
models in terms of how well they can generalize to changes in the data distribution.
We first study a method for co-training sparse and dense models, at a lower cost
compared to regular training. With our method we can obtain very accurate sparse
networks, and dense models that can recover the baseline accuracy. Furthermore,
we are able to more easily analyze the differences, at prediction level, between
the sparse-dense model pairs. Next, we investigate the generalization properties
of sparse neural networks in more detail, by studying how well different sparse
models trained on a larger task can adapt to smaller, more specialized tasks,
in a transfer learning scenario. Our analysis across multiple pruning methods
and sparsity levels reveals that sparse models provide features that can transfer
similarly to or better than the dense baseline. However, the choice of the pruning
method plays an important role, and can influence the results when the features
are fixed (linear finetuning), or when they are allowed to adapt to the new task
(full finetuning). Using sparse models with fixed masks for finetuning on new
tasks has an important practical advantage, as it enables training neural networks
on smaller devices. However, one drawback of current pruning methods is that the
entire training cycle has to be repeated to obtain the initial sparse model, for
every sparsity target; in consequence, the entire training process is costly and
also multiple models need to be stored. In the last part of the thesis we propose
a method that can train accurate dense models that are compressible in a single
step, to multiple sparsity levels, without additional finetuning. Our method results
in sparse models that can be competitive with existing pruning methods, and which
can also successfully generalize to new tasks."
acknowledged_ssus:
- _id: ScienComp
alternative_title:
- ISTA Thesis
article_processing_charge: No
author:
- first_name: Elena-Alexandra
full_name: Peste, Elena-Alexandra
id: 32D78294-F248-11E8-B48F-1D18A9856A87
last_name: Peste
citation:
ama: Krumes A. Efficiency and generalization of sparse neural networks. 2023. doi:10.15479/at:ista:13074
apa: Krumes, A. (2023). Efficiency and generalization of sparse neural networks.
Institute of Science and Technology Austria. https://doi.org/10.15479/at:ista:13074
chicago: Krumes, Alexandra. “Efficiency and Generalization of Sparse Neural Networks.”
Institute of Science and Technology Austria, 2023. https://doi.org/10.15479/at:ista:13074.
ieee: A. Krumes, “Efficiency and generalization of sparse neural networks,” Institute
of Science and Technology Austria, 2023.
ista: Krumes A. 2023. Efficiency and generalization of sparse neural networks. Institute
of Science and Technology Austria.
mla: Krumes, Alexandra. Efficiency and Generalization of Sparse Neural Networks.
Institute of Science and Technology Austria, 2023, doi:10.15479/at:ista:13074.
short: A. Krumes, Efficiency and Generalization of Sparse Neural Networks, Institute
of Science and Technology Austria, 2023.
corr_author: '1'
date_created: 2023-05-23T17:07:53Z
date_published: 2023-05-23T00:00:00Z
date_updated: 2026-04-07T13:30:20Z
day: '23'
ddc:
- '000'
degree_awarded: PhD
department:
- _id: GradSch
- _id: DaAl
- _id: ChLa
doi: 10.15479/at:ista:13074
ec_funded: 1
file:
- access_level: open_access
checksum: 6b3354968403cb9d48cc5a83611fb571
content_type: application/pdf
creator: epeste
date_created: 2023-05-24T16:11:16Z
date_updated: 2023-05-24T16:11:16Z
file_id: '13087'
file_name: PhD_Thesis_Alexandra_Peste_final.pdf
file_size: 2152072
relation: main_file
success: 1
- access_level: closed
checksum: 8d0df94bbcf4db72c991f22503b3fd60
content_type: application/zip
creator: epeste
date_created: 2023-05-24T16:12:59Z
date_updated: 2023-05-24T16:12:59Z
file_id: '13088'
file_name: PhD_Thesis_APeste.zip
file_size: 1658293
relation: source_file
file_date_updated: 2023-05-24T16:12:59Z
has_accepted_license: '1'
language:
- iso: eng
month: '05'
oa: 1
oa_version: Published Version
page: '147'
project:
- _id: 2564DBCA-B435-11E9-9278-68D0E5697425
call_identifier: H2020
grant_number: '665385'
name: International IST Doctoral Program
- _id: 268A44D6-B435-11E9-9278-68D0E5697425
call_identifier: H2020
grant_number: '805223'
name: Elastic Coordination for Scalable Machine Learning
publication_identifier:
issn:
- 2663-337X
publication_status: published
publisher: Institute of Science and Technology Austria
related_material:
record:
- id: '13053'
relation: part_of_dissertation
status: public
- id: '11458'
relation: part_of_dissertation
status: public
- id: '12299'
relation: part_of_dissertation
status: public
status: public
supervisor:
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
- first_name: Dan-Adrian
full_name: Alistarh, Dan-Adrian
id: 4A899BFC-F248-11E8-B48F-1D18A9856A87
last_name: Alistarh
orcid: 0000-0003-3650-940X
title: Efficiency and generalization of sparse neural networks
type: dissertation
user_id: ba8df636-2132-11f1-aed0-ed93e2281fdd
year: '2023'
...
---
_id: '13053'
abstract:
- lang: eng
text: 'Deep neural networks (DNNs) often have to be compressed, via pruning and/or
quantization, before they can be deployed in practical settings. In this work
we propose a new compression-aware minimizer dubbed CrAM that modifies the optimization
step in a principled way, in order to produce models whose local loss behavior
is stable under compression operations such as pruning. Thus, dense models trained
via CrAM should be compressible post-training, in a single step, without significant
accuracy loss. Experimental results on standard benchmarks, such as residual networks
for ImageNet classification and BERT models for language modelling, show that
CrAM produces dense models that can be more accurate than the standard SGD/Adam-based
baselines, but which are stable under weight pruning: specifically, we can prune
models in one-shot to 70-80% sparsity with almost no accuracy loss, and to 90%
with reasonable (∼1%) accuracy loss, which is competitive with gradual compression
methods. Additionally, CrAM can produce sparse models which perform well for transfer
learning, and it also works for semi-structured 2:4 pruning patterns supported
by GPU hardware. The code for reproducing the results is available at this https
URL .'
acknowledged_ssus:
- _id: ScienComp
acknowledgement: "AP, EK, DA received funding from the European Research Council (ERC)
under the European\r\nUnion’s Horizon 2020 research and innovation programme (grant
agreement No 805223 ScaleML). AV acknowledges the support of the French Agence Nationale
de la Recherche (ANR), under grant ANR-21-CE48-0016 (project COMCOPT). We further
acknowledge the support from the Scientific Service Units (SSU) of ISTA through
resources provided by Scientific Computing (SciComp)."
article_processing_charge: No
arxiv: 1
author:
- first_name: Elena-Alexandra
full_name: Peste, Elena-Alexandra
id: 32D78294-F248-11E8-B48F-1D18A9856A87
last_name: Peste
- first_name: Adrian
full_name: Vladu, Adrian
last_name: Vladu
- first_name: Eldar
full_name: Kurtic, Eldar
id: 47beb3a5-07b5-11eb-9b87-b108ec578218
last_name: Kurtic
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
- first_name: Dan-Adrian
full_name: Alistarh, Dan-Adrian
id: 4A899BFC-F248-11E8-B48F-1D18A9856A87
last_name: Alistarh
orcid: 0000-0003-3650-940X
citation:
ama: 'Krumes A, Vladu A, Kurtic E, Lampert C, Alistarh D-A. CrAM: A Compression-Aware
Minimizer. In: 11th International Conference on Learning Representations .
OpenReview; 2023.'
apa: 'Krumes, A., Vladu, A., Kurtic, E., Lampert, C., & Alistarh, D.-A. (2023).
CrAM: A Compression-Aware Minimizer. In 11th International Conference on Learning
Representations . Kigali, Rwanda : OpenReview.'
chicago: 'Krumes, Alexandra, Adrian Vladu, Eldar Kurtic, Christoph Lampert, and
Dan-Adrian Alistarh. “CrAM: A Compression-Aware Minimizer.” In 11th International
Conference on Learning Representations . OpenReview, 2023.'
ieee: 'A. Krumes, A. Vladu, E. Kurtic, C. Lampert, and D.-A. Alistarh, “CrAM: A
Compression-Aware Minimizer,” in 11th International Conference on Learning
Representations , Kigali, Rwanda , 2023.'
ista: 'Krumes A, Vladu A, Kurtic E, Lampert C, Alistarh D-A. 2023. CrAM: A Compression-Aware
Minimizer. 11th International Conference on Learning Representations . ICLR: International
Conference on Learning Representations.'
mla: 'Krumes, Alexandra, et al. “CrAM: A Compression-Aware Minimizer.” 11th International
Conference on Learning Representations , OpenReview, 2023.'
short: A. Krumes, A. Vladu, E. Kurtic, C. Lampert, D.-A. Alistarh, in:, 11th International
Conference on Learning Representations , OpenReview, 2023.
conference:
end_date: 2023-05-05
location: 'Kigali, Rwanda '
name: 'ICLR: International Conference on Learning Representations'
start_date: 2023-05-01
corr_author: '1'
date_created: 2023-05-23T11:36:18Z
date_published: 2023-05-01T00:00:00Z
date_updated: 2026-04-07T13:30:19Z
day: '01'
ddc:
- '000'
department:
- _id: GradSch
- _id: DaAl
- _id: ChLa
ec_funded: 1
external_id:
arxiv:
- '2207.14200'
file:
- access_level: open_access
checksum: a6eec897e13a91cdc3eeaf309801752c
content_type: application/pdf
creator: dernst
date_created: 2024-07-22T09:09:45Z
date_updated: 2024-07-22T09:09:45Z
file_id: '17294'
file_name: 2023_ICLR_Peste.pdf
file_size: 458201
relation: main_file
success: 1
file_date_updated: 2024-07-22T09:09:45Z
has_accepted_license: '1'
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://openreview.net/pdf?id=_eTZBs-yedr
month: '05'
oa: 1
oa_version: Published Version
project:
- _id: 268A44D6-B435-11E9-9278-68D0E5697425
call_identifier: H2020
grant_number: '805223'
name: Elastic Coordination for Scalable Machine Learning
publication: '11th International Conference on Learning Representations '
publication_status: published
publisher: OpenReview
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://github.com/IST-DASLab/CrAM
record:
- id: '13074'
relation: dissertation_contains
status: public
status: public
title: 'CrAM: A Compression-Aware Minimizer'
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2023'
...
---
_id: '14771'
abstract:
- lang: eng
text: Pruning—that is, setting a significant subset of the parameters of a neural
network to zero—is one of the most popular methods of model compression. Yet,
several recent works have raised the issue that pruning may induce or exacerbate
bias in the output of the compressed model. Despite existing evidence for this
phenomenon, the relationship between neural network pruning and induced bias is
not well-understood. In this work, we systematically investigate and characterize
this phenomenon in Convolutional Neural Networks for computer vision. First, we
show that it is in fact possible to obtain highly-sparse models, e.g. with less
than 10% remaining weights, which do not decrease in accuracy nor substantially
increase in bias when compared to dense models. At the same time, we also find
that, at higher sparsities, pruned models exhibit higher uncertainty in their
outputs, as well as increased correlations, which we directly link to increased
bias. We propose easy-to-use criteria which, based only on the uncompressed model,
establish whether bias will increase with pruning, and identify the samples most
susceptible to biased predictions post-compression. Our code can be found at https://github.com/IST-DASLab/pruned-vision-model-bias.
acknowledgement: The authors would like to sincerely thank Sara Hooker for her feedback
during the development of this work. EI was supported in part by the FWF DK VGSCO,
grant agreement number W1260-N35. AP and DA acknowledge generous ERC support, via
Starting Grant 805223 ScaleML.
article_processing_charge: No
arxiv: 1
author:
- first_name: Eugenia B
full_name: Iofinova, Eugenia B
id: f9a17499-f6e0-11ea-865d-fdf9a3f77117
last_name: Iofinova
orcid: 0000-0002-7778-3221
- first_name: Elena-Alexandra
full_name: Peste, Elena-Alexandra
id: 32D78294-F248-11E8-B48F-1D18A9856A87
last_name: Peste
- first_name: Dan-Adrian
full_name: Alistarh, Dan-Adrian
id: 4A899BFC-F248-11E8-B48F-1D18A9856A87
last_name: Alistarh
orcid: 0000-0003-3650-940X
citation:
ama: 'Iofinova EB, Krumes A, Alistarh D-A. Bias in pruned vision models: In-depth
analysis and countermeasures. In: 2023 IEEE/CVF Conference on Computer Vision
and Pattern Recognition. IEEE; 2023:24364-24373. doi:10.1109/cvpr52729.2023.02334'
apa: 'Iofinova, E. B., Krumes, A., & Alistarh, D.-A. (2023). Bias in pruned
vision models: In-depth analysis and countermeasures. In 2023 IEEE/CVF Conference
on Computer Vision and Pattern Recognition (pp. 24364–24373). Vancouver, BC,
Canada: IEEE. https://doi.org/10.1109/cvpr52729.2023.02334'
chicago: 'Iofinova, Eugenia B, Alexandra Krumes, and Dan-Adrian Alistarh. “Bias
in Pruned Vision Models: In-Depth Analysis and Countermeasures.” In 2023 IEEE/CVF
Conference on Computer Vision and Pattern Recognition, 24364–73. IEEE, 2023.
https://doi.org/10.1109/cvpr52729.2023.02334.'
ieee: 'E. B. Iofinova, A. Krumes, and D.-A. Alistarh, “Bias in pruned vision models:
In-depth analysis and countermeasures,” in 2023 IEEE/CVF Conference on Computer
Vision and Pattern Recognition, Vancouver, BC, Canada, 2023, pp. 24364–24373.'
ista: 'Iofinova EB, Krumes A, Alistarh D-A. 2023. Bias in pruned vision models:
In-depth analysis and countermeasures. 2023 IEEE/CVF Conference on Computer Vision
and Pattern Recognition. CVPR: Conference on Computer Vision and Pattern Recognition,
24364–24373.'
mla: 'Iofinova, Eugenia B., et al. “Bias in Pruned Vision Models: In-Depth Analysis
and Countermeasures.” 2023 IEEE/CVF Conference on Computer Vision and Pattern
Recognition, IEEE, 2023, pp. 24364–73, doi:10.1109/cvpr52729.2023.02334.'
short: E.B. Iofinova, A. Krumes, D.-A. Alistarh, in:, 2023 IEEE/CVF Conference on
Computer Vision and Pattern Recognition, IEEE, 2023, pp. 24364–24373.
conference:
end_date: 2023-06-24
location: Vancouver, BC, Canada
name: 'CVPR: Conference on Computer Vision and Pattern Recognition'
start_date: 2023-06-17
corr_author: '1'
date_created: 2024-01-10T08:42:40Z
date_published: 2023-08-22T00:00:00Z
date_updated: 2026-05-19T11:20:27Z
day: '22'
department:
- _id: DaAl
- _id: ChLa
doi: 10.1109/cvpr52729.2023.02334
ec_funded: 1
external_id:
arxiv:
- '2304.12622'
isi:
- '001062531308068'
isi: 1
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2304.12622
month: '08'
oa: 1
oa_version: Preprint
page: 24364-24373
project:
- _id: 9B9290DE-BA93-11EA-9121-9846C619BF3A
grant_number: W1260-N35
name: Vienna Graduate School on Computational Optimization
- _id: 268A44D6-B435-11E9-9278-68D0E5697425
call_identifier: H2020
grant_number: '805223'
name: Elastic Coordination for Scalable Machine Learning
publication: 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition
publication_identifier:
eisbn:
- '9798350301298'
eissn:
- 2575-7075
publication_status: published
publisher: IEEE
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://github.com/IST-DASLab/pruned-vision-model-bias
record:
- id: '21854'
relation: dissertation_contains
status: public
status: public
title: 'Bias in pruned vision models: In-depth analysis and countermeasures'
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2023'
...