Zobrazeno 1 - 10
of 27
pro vyhledávání: '"Jin, Charles"'
Autor:
Jin, Charles, Rinard, Martin
As language models (LMs) deliver increasing performance on a range of NLP tasks, probing classifiers have become an indispensable technique in the effort to better understand their inner workings. A typical setup involves (1) defining an auxiliary ta
Externí odkaz:
http://arxiv.org/abs/2407.13765
Autor:
Jin, Charles, Rinard, Martin
Publikováno v:
PMLR 235:22160-22184, 2024
We present evidence that language models (LMs) of code can learn to represent the formal semantics of programs, despite being trained only to perform next-token prediction. Specifically, we train a Transformer model on a synthetic corpus of programs
Externí odkaz:
http://arxiv.org/abs/2305.11169
This work studies the problem of ad hoc teamwork in teams composed of agents with differing computational capabilities. We consider cooperative multi-player games in which each agent's policy is constrained by a private capability parameter, and agen
Externí odkaz:
http://arxiv.org/abs/2304.13957
Publikováno v:
Proc. ACM Program. Lang., Vol. 6, No. OOPSLA2, Article 166. Publication date: October 2022
In the past few years, neural architecture search (NAS) has become an increasingly important tool within the deep learning community. Despite the many recent successes of NAS, however, most existing approaches operate within highly structured design
Externí odkaz:
http://arxiv.org/abs/2205.03960
We propose a novel clustering mechanism based on an incompatibility property between subsets of data that emerges during model training. This mechanism partitions the dataset into subsets that generalize only to themselves, i.e., training on one subs
Externí odkaz:
http://arxiv.org/abs/2105.03692
Autor:
Jin, Charles, Rinard, Martin
We propose a novel setting for learning, where the input domain is the image of a map defined on the product of two sets, one of which completely determines the labels. We derive a new risk bound for this setting that decomposes into a bias and an er
Externí odkaz:
http://arxiv.org/abs/2005.14707
Autor:
Jin, Charles, Rinard, Martin
We apply concepts from manifold regularization to develop new regularization techniques for training locally stable deep neural networks. Our regularizers are based on a sparsification of the graph Laplacian which holds with high probability when the
Externí odkaz:
http://arxiv.org/abs/2003.04286
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.
Autor:
Jin, Charles, Rinard, Martin
We present evidence that language models can learn meaning despite being trained only to perform next token prediction on text, specifically a corpus of programs. Each program is preceded by a specification in the form of (textual) input-output examp
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6e47dfd2865fc6e21e8c190f8b81ac89
5.37 ASSOCIATIONS OF SCREEN TIME AND DEPRESSION IN CHILDREN IN A US NATIONALLY REPRESENTATIVE SAMPLE
Publikováno v:
In Journal of the American Academy of Child & Adolescent Psychiatry October 2019 58(10) Supplement:S257-S257