# Re-ReST: Reflection-Reinforced Self-Training for Language Agents

Source: [Re-ReST: Reflection-Reinforced Self-Training for Language Agents](<https://www.sota2.com/research/paper/68fa78621e209131cea7816d>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T04:37:17.197Z

Authors: Zi-Yi Dou, Cheng-Fu Yang, Xueqing Wu, Kai-Wei Chang, Nanyun Peng

Published: 2024-06-03T16:21:38.000Z

## Abstract

Finetuning language agents with reasoning-action trajectories is effective, but obtaining these trajectories from human annotations or stronger models is costly and sometimes impractical. In this paper, we investigate the use of self-training in language agents, which can generate supervision from the agent itself, offering a promising alternative without relying on human or stronger model demonstrations. Self-training, however, requires high-quality model-generated samples, which are hard to obtain for challenging language agent tasks. To address this, we present Reflection-Reinforced Self-Training (Re-ReST), which uses a \\textit{reflector} to refine low-quality generated samples during self-training. The reflector takes the agent's output and feedback from an external environment (e.g., unit test results in code generation) to produce improved samples. This technique enhances the quality of inferior samples and efficiently enriches the self-training dataset with higher-quality samples. We conduct extensive experiments on open-source language agents across tasks, including multi-hop question answering, sequential decision-making, code generation, visual question answering, and text-to-image generation. The results demonstrate the effectiveness of self-training and Re-ReST in language agent tasks, with self-training improving baselines by 7.6\\% on HotpotQA and 28.4\\% on AlfWorld, and Re-ReST further boosting performance by 2.0\\% and 14.1\\%, respectively. Our studies also confirm the efficiency of using a reflector to generate high-quality samples for self-training. Moreover, we demonstrate a method to employ reflection during inference without ground-truth feedback, addressing the limitation of previous reflection work. Our code is released at https://github.com/PlusLabNLP/Re-ReST.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Mathematical Reasoning on Minerva Math](<https://www.sota2.com/research/sota/mathematical-reasoning-on-minerva-math.md>): Mathematical Reasoning; Minerva Math; Accuracy: 29.04.

- [Mathematical Reasoning on AIME 24/25 (Accuracy)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-aime-24-25-accuracy.md>): Mathematical Reasoning; AIME 24/25; Accuracy: 5.

- [Mathematical Reasoning on Olympiad Bench (Accuracy)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-olympiad-bench-accuracy.md>): Mathematical Reasoning; Olympiad Bench; Accuracy: 15.56.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2406.01495>)

[PDF](<https://arxiv.org/pdf/2406.01495.pdf>)
