Work

RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Horizon Generation

Year: 2024

Type: preprint

Abstract: We explore how iterative revising a chain of thoughts with the help of information retrieval significantly improves large language models' reasoning and generation ability in long-horizon generation tasks, while hugely mitigating hallucination. In particular, the proposed method -- *retrieval-augmented thoughts* (RAT) -- revises each thought step one by one with retrieved information relevant to the task query, the current and the past thought steps, after the initial zero-shot CoT is generated. Applying RAT to GPT-3.5, GPT-4, and CodeLLaMA-7b substantially improves their performances on various long-horizon generation tasks; on average of relatively increasing rating scores by 13.63% on code generation, 16.96% on mathematical reasoning, 19.2% on creative writing, and 42.78% on embodied task planning. The demo page can be found at https://craftjarvis.github.io/RAT

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.2403.05313

Authors: Wang, Zihao, Liu, Anji, Lin, Haowei, Li, Jiaqi, Ma, Xiaojian, Liang, Yitao

Institutions: Peking University, University of California, Los Angeles, Beijing Institute for General Artificial Intelligence

Language: English

Topic: Topic Modeling

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: National Science Foundation, Tencent

Awards: Collaborative Research: RI: AF: Medium: Exchanging Knowledge Beyond Data Between Human and Machine Learner, CAREER: Towards a New Synthesis of Statistical Learning and Logical Reasoning, FMitF: Opening Up the Black Box of Probabilistic Program Inference

OpenAlex ID: W4392678293

Cites:

Cited by: 11

Related to: 10