Nur Lan
@nurikolan.bsky.social
Reposted by Nur Lan
📄 New paper: “A Minimum Description Length Approach to Regularization in Neural Networks”
with Orr Well, Emmanuel Chemla, @rkatzir.bsky.social and @nurikolan.bsky.social
We explore why neural networks often struggle with simple, structured tasks.
Spoiler: our regularizers might be the problem.
🧵
with Orr Well, Emmanuel Chemla, @rkatzir.bsky.social and @nurikolan.bsky.social
We explore why neural networks often struggle with simple, structured tasks.
Spoiler: our regularizers might be the problem.
🧵
May 24, 2025 at 4:01 PM
📄 New paper: “A Minimum Description Length Approach to Regularization in Neural Networks”
with Orr Well, Emmanuel Chemla, @rkatzir.bsky.social and @nurikolan.bsky.social
We explore why neural networks often struggle with simple, structured tasks.
Spoiler: our regularizers might be the problem.
🧵
with Orr Well, Emmanuel Chemla, @rkatzir.bsky.social and @nurikolan.bsky.social
We explore why neural networks often struggle with simple, structured tasks.
Spoiler: our regularizers might be the problem.
🧵
Reposted by Nur Lan
LLMs are bad theories of human linguistic cognition but can be useful proxies for less bad ones. Here's how and why it might be useful:
arxiv.org/abs/2502.07687
Joint paper with @nurikolan.bsky.social, Emmanuel Chemla and @rkatzir.bsky.social
arxiv.org/abs/2502.07687
Joint paper with @nurikolan.bsky.social, Emmanuel Chemla and @rkatzir.bsky.social
Large Language Models as Proxies for Theories of Human Linguistic Cognition
We consider the possible role of current large language models (LLMs) in the study of human linguistic cognition. We focus on the use of such models as proxies for theories of cognition that are relat...
arxiv.org
February 12, 2025 at 10:40 PM
LLMs are bad theories of human linguistic cognition but can be useful proxies for less bad ones. Here's how and why it might be useful:
arxiv.org/abs/2502.07687
Joint paper with @nurikolan.bsky.social, Emmanuel Chemla and @rkatzir.bsky.social
arxiv.org/abs/2502.07687
Joint paper with @nurikolan.bsky.social, Emmanuel Chemla and @rkatzir.bsky.social
⭐️🗞️ Accepted to ACL 2024 main conference! #ACL2024NLP
Neural nets can in theory learn formal languages such as aⁿbⁿ & Dyck. Yet no one ever finds such nets using standard techniques. Why?
We suggest that the culprit might have been the objective function all along 👇
arxiv.org/abs/2402.10013
Neural nets can in theory learn formal languages such as aⁿbⁿ & Dyck. Yet no one ever finds such nets using standard techniques. Why?
We suggest that the culprit might have been the objective function all along 👇
arxiv.org/abs/2402.10013
🧪🗞️ New paper with Emmanuel Chemla and @rkatzir.bsky.social:
Neural nets offer good approximation but consistently fail to generalize perfectly, even when perfect solutions are proved to exist.
We check whether the culprit might be their training objective.
arxiv.org/abs/2402.10013
Neural nets offer good approximation but consistently fail to generalize perfectly, even when perfect solutions are proved to exist.
We check whether the culprit might be their training objective.
arxiv.org/abs/2402.10013
June 17, 2024 at 6:54 PM
⭐️🗞️ Accepted to ACL 2024 main conference! #ACL2024NLP
Neural nets can in theory learn formal languages such as aⁿbⁿ & Dyck. Yet no one ever finds such nets using standard techniques. Why?
We suggest that the culprit might have been the objective function all along 👇
arxiv.org/abs/2402.10013
Neural nets can in theory learn formal languages such as aⁿbⁿ & Dyck. Yet no one ever finds such nets using standard techniques. Why?
We suggest that the culprit might have been the objective function all along 👇
arxiv.org/abs/2402.10013
🧪🗞️ New paper with Emmanuel Chemla and @rkatzir.bsky.social:
Neural nets offer good approximation but consistently fail to generalize perfectly, even when perfect solutions are proved to exist.
We check whether the culprit might be their training objective.
arxiv.org/abs/2402.10013
Neural nets offer good approximation but consistently fail to generalize perfectly, even when perfect solutions are proved to exist.
We check whether the culprit might be their training objective.
arxiv.org/abs/2402.10013
February 17, 2024 at 6:13 PM
🧪🗞️ New paper with Emmanuel Chemla and @rkatzir.bsky.social:
Neural nets offer good approximation but consistently fail to generalize perfectly, even when perfect solutions are proved to exist.
We check whether the culprit might be their training objective.
arxiv.org/abs/2402.10013
Neural nets offer good approximation but consistently fail to generalize perfectly, even when perfect solutions are proved to exist.
We check whether the culprit might be their training objective.
arxiv.org/abs/2402.10013
Douglas Hofstadter on toy tasks, in Waking Up from the Boolean Dream, 1982
November 23, 2023 at 10:09 AM
Douglas Hofstadter on toy tasks, in Waking Up from the Boolean Dream, 1982
⚡ 🗞️ New up-to-date version of Large Language Models and the Argument from the Poverty of the Stimulus, work with Emmanuel Chemla and @rkatzir.bsky.social:
ling.auf.net/lingbuzz/006...
ling.auf.net/lingbuzz/006...
November 21, 2023 at 4:14 PM
⚡ 🗞️ New up-to-date version of Large Language Models and the Argument from the Poverty of the Stimulus, work with Emmanuel Chemla and @rkatzir.bsky.social:
ling.auf.net/lingbuzz/006...
ling.auf.net/lingbuzz/006...
How well can neural networks generalize from how little data?
New work with Emmanuel Chemla and Roni Katzir:
Benchmark:
github.com/taucompling/...
Paper:
aclanthology.org/2023.clasp-1...
🧵
New work with Emmanuel Chemla and Roni Katzir:
Benchmark:
github.com/taucompling/...
Paper:
aclanthology.org/2023.clasp-1...
🧵
GitHub - taucompling/bliss: 🧘 BLISS – a Benchmark for Language Induction from Small Sets
🧘 BLISS – a Benchmark for Language Induction from Small Sets - GitHub - taucompling/bliss: 🧘 BLISS – a Benchmark for Language Induction from Small Sets
github.com
October 2, 2023 at 9:38 AM
How well can neural networks generalize from how little data?
New work with Emmanuel Chemla and Roni Katzir:
Benchmark:
github.com/taucompling/...
Paper:
aclanthology.org/2023.clasp-1...
🧵
New work with Emmanuel Chemla and Roni Katzir:
Benchmark:
github.com/taucompling/...
Paper:
aclanthology.org/2023.clasp-1...
🧵