Zobrazeno 1 - 10
of 38
pro vyhledávání: '"Chen, Angelica"'
Autor:
Chen, Angelica, Stanton, Samuel D., Alberstein, Robert G., Watkins, Andrew M., Bonneau, Richard, Gligorijevi, Vladimir, Cho, Kyunghyun, Frey, Nathan C.
Large language models (LLMs) have recently shown significant potential in various biological tasks such as protein engineering and molecule design. These tasks typically involve black-box discrete sequence optimization, where the challenge lies in ge
Externí odkaz:
http://arxiv.org/abs/2410.22296
Autor:
Chen, Angelica, Malladi, Sadhika, Zhang, Lily H., Chen, Xinyi, Zhang, Qiuyi, Ranganath, Rajesh, Cho, Kyunghyun
Preference learning algorithms (e.g., RLHF and DPO) are frequently used to steer LLMs to produce generations that are more preferred by humans, but our understanding of their inner workings is still limited. In this work, we study the conventional wi
Externí odkaz:
http://arxiv.org/abs/2405.19534
We consider regret minimization in repeated games with a very large number of actions. Such games are inherent in the setting of AI Safety via Debate \cite{irving2018ai}, and more generally games whose actions are language-based. Existing algorithms
Externí odkaz:
http://arxiv.org/abs/2312.04792
Most interpretability research in NLP focuses on understanding the behavior and features of a fully trained model. However, certain insights into model behavior may only be accessible by observing the trajectory of the training process. We present a
Externí odkaz:
http://arxiv.org/abs/2309.07311
The impact of randomness on model training is poorly understood. How do differences in data order and initialization actually manifest in the model, such that some training runs outperform others or converge faster? Furthermore, how can we interpret
Externí odkaz:
http://arxiv.org/abs/2308.09543
Autor:
Chen, Angelica, Phang, Jason, Parrish, Alicia, Padmakumar, Vishakh, Zhao, Chen, Bowman, Samuel R., Cho, Kyunghyun
Publikováno v:
Transactions on Machine Learning Research (2024)
Large language models (LLMs) have achieved widespread success on a variety of in-context few-shot tasks, but this success is typically evaluated via correctness rather than consistency. We argue that self-consistency is an important criteria for vali
Externí odkaz:
http://arxiv.org/abs/2305.14279
Autor:
Scheurer, Jérémy, Campos, Jon Ander, Korbak, Tomasz, Chan, Jun Shern, Chen, Angelica, Cho, Kyunghyun, Perez, Ethan
Pretrained language models often generate outputs that are not in line with human preferences, such as harmful text or factually incorrect summaries. Recent work approaches the above issues by learning from a simple form of human feedback: comparison
Externí odkaz:
http://arxiv.org/abs/2303.16755
Autor:
Chen, Angelica, Scheurer, Jérémy, Korbak, Tomasz, Campos, Jon Ander, Chan, Jun Shern, Bowman, Samuel R., Cho, Kyunghyun, Perez, Ethan
The potential for pre-trained large language models (LLMs) to use natural language feedback at inference time has been an exciting recent development. We build upon this observation by formalizing an algorithm for learning from natural language feedb
Externí odkaz:
http://arxiv.org/abs/2303.16749
Given the recent impressive accomplishments of language models (LMs) for code generation, we explore the use of LMs as adaptive mutation and crossover operators for an evolutionary neural architecture search (NAS) algorithm. While NAS still proves to
Externí odkaz:
http://arxiv.org/abs/2302.14838
Autor:
Korbak, Tomasz, Shi, Kejian, Chen, Angelica, Bhalerao, Rasika, Buckley, Christopher L., Phang, Jason, Bowman, Samuel R., Perez, Ethan
Language models (LMs) are pretrained to imitate internet text, including content that would violate human preferences if generated by an LM: falsehoods, offensive comments, personally identifiable information, low-quality or buggy code, and more. Her
Externí odkaz:
http://arxiv.org/abs/2302.08582