Dani's Stack140 ReAct Ⅱ 이번 글은 지난 ReAct 글을 좀 더 자세하게 정리한 글입니다. https://hexists.tistory.com/252 ReAct ReAct: Synergizing Reasoning and Acting in Language Models https://ai.googleblog.com/2022/11/react-synergizing-reasoning-and-acting.html ReAct: Synergizing Reasoning and Acting in Language Models Posted by Shunyu Yao, Student Researcher, and Yuan Cao, Rese hexists.tistory.com 논문을 단락별로 3줄 요약 형식으로 정리해봅니다. DeepL을 사용해서 번역했습니다.. 2023. 6. 16. ReAct ReAct: Synergizing Reasoning and Acting in Language Models https://ai.googleblog.com/2022/11/react-synergizing-reasoning-and-acting.html ReAct: Synergizing Reasoning and Acting in Language Models Posted by Shunyu Yao, Student Researcher, and Yuan Cao, Research Scientist, Google Research, Brain Team --> Recent advances have expanded the applicability of language models (LM) to downstream tasks. O.. 2023. 5. 31. MLM vs CLM Maksed Language Model(MLM) vs Casual Language Model(CLM) https://towardsdatascience.com/understanding-masked-language-models-mlm-and-causal-language-models-clm-in-nlp-194c15f56a5 Understanding Masked Language Models (MLM) and Causal Language Models (CLM) in NLP Language Models in NLP (Visuals and Examples) towardsdatascience.com 위 링크를 바탕으로 정리한 내용입니다. 1. Maksed Language Model(MLM) - bidirectional.. 2023. 4. 20. InstructGPT Evaluation https://openai.com/research/instruction-following https://arxiv.org/abs/2203.02155 instructGPT의 Evaluation에 대해 정리합니다. 위 논문에서 3.6 Evaluation 부분을 보고 이해한대로 정리한 내용입니다. 잘못 이해한 부분이 있다면 언제든지 알려주세요. Alignment 평가에서 중요한 개념 Leike et al. (2018): user intentions에 따라서 행동하는 모델을 훈련 Askell et al. (2021): helpful, honest, and harmless 되도록 정렬 instructGPT에서는 Askell과 유사한 framework를 사용 helpful 평가 모델은 1) instructions.. 2023. 4. 17. 이전 1 2 3 4 5 ··· 35 다음
