Observation steps. Thought can reason about the current situation, and Action can be three types:<Actionの説明> Here are some examples. <例を示す> Chain of Thought(Wei et al., 2022) と Few-Shot(Brown et al., 2020)を組み合わせたプロン プト
Mukai. Misreading Chat: #143 Can Language Models Resolve Real-World GitHub Issues?. ポッドキャ スト, 2024 • Vaswani, A. et al. 2017. Attention is All You Need. • Radford, A. et al. 2018. Improving Language Understanding by Generative Pre-Training.(GPT)
with human feedback. • Chen, M. et al. 2021. Evaluating Large Language Models Trained on Code.(HumanEval / Codex) • Hendrycks, D. et al. 2021. Measuring Coding Challenge Competence with APPS. • Ouyang, L. et al. 2022. Training language models to follow instructions with human feedback.(InstructGPT)
Reasoning in Large Language Models. • Yao, S. et al. 2022. ReAct: Synergizing Reasoning and Acting in Language Models. • Liu, H. et al. 2023. Lost in the Middle: How Language Models Use Long Context. • Shinn, N. et al. 2023. Reflexion: Language Agents with Verbal Reinforcement Learning.