<CoT, Prompting> [Google DeepMind] Chain-of-Thought Reasoning Without Prompting (2024.02)

<CoT, Prompting> [Google DeepMind] Chain-of-Thought Reasoning Without Prompting (2024.02)

관심 있는 NLP 논문을 읽어보고 간단히 정리했습니다. 혹시 부족하거나 잘못된 내용이 있다면 댓글 부탁드립니다 ️ usechatgpt init success [Google DeepMind] - LLM의 decoding process를 변경함으로써 prompting 없이 CoT reasoning paths를 유도할 수 있다고 주장 - top-k개의 alternative tokens를 조사하여 sequences에 내재하는 CoT paths를 확인. 즉 LLM에게 intrinsic reasoning ability가 있다고 주장 1. Introduction LLM이 눈부신 발전을 거듭하는 과정에는 reasoning task에 대한 수행 능력의 발전이 큰 몫을 차지하고 있습니다. 여기에는 특히나 CoT (C..

원문링크 : <CoT, Prompting> [Google DeepMind] Chain-of-Thought Reasoning Without Prompting (2024.02)

 등록된 다른 글

Transpose Convolutions, U-Net Architecture Intuition

Transpose Convolutions, U-Net Architecture Intuition

3.3. LU 분해

3.3. LU 분해

<Decompose> Self-Discover: Large Language Models Self-Compose Reasoning Structures (2024.02)

<Decompose> Self-Discover: Large Language Models Self-Compose Reasoning Structures (2024.02)

Introduction to Deep Learning

Introduction to Deep Learning

[BOJ] 1260 : DFS와 BFS [그래프 이론](Python)

[BOJ] 1260 : DFS와 BFS [그래프 이론](Python)

<CoT> [CoVe] Chain-of-Verification Reduces Hallucination in Large Language Models

<CoT> [CoVe] Chain-of-Verification Reduces Hallucination in Large Language Models

인공지능 최신 논문/뉴스 follow-up 꿀팁 대공개!! (NLP, LLM 위주 )

인공지능 최신 논문/뉴스 follow-up 꿀팁 대공개!! (NLP, LLM 위주 )

<Retrieval, In-Context Learning> RAVEN: In-Context Learning with Retrieval Augmented Encoder-Decoder Language Models

<Retrieval, In-Context Learning> RAVEN: In-Context Learning with Retrieval Augmented Encoder-Decoder Language Models