Exploring Expert Failures Improves LLM Agent Tuning

• URL: http://arxiv.org/abs/2504.13145v2
• Date: Fri, 18 Apr 2025 19:36:21 GMT
• Title: Exploring Expert Failures Improves LLM Agent Tuning
• Authors: Li-Cheng Lan, Andrew Bai, Minhao Cheng, Cho-Jui Hsieh, Tianyi Zhou,
• Abstract summary: We propose Exploring Expert Failures (EEF), which identifies beneficial actions from failed expert trajectories.<n>EEF successfully solves some previously unsolvable subtasks and improves agent tuning performance.
• Score: 74.0772570556016
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Language Models (LLMs) have shown tremendous potential as agents, excelling at tasks that require multiple rounds of reasoning and interactions. Rejection Sampling Fine-Tuning (RFT) has emerged as an effective method for finetuning LLMs as agents: it first imitates expert-generated successful trajectories and further improves agentic skills through iterative fine-tuning on successful, self-generated trajectories. However, since the expert (e.g., GPT-4) succeeds primarily on simpler subtasks and RFT inherently favors simpler scenarios, many complex subtasks remain unsolved and persistently out-of-distribution (OOD). Upon investigating these challenging subtasks, we discovered that previously failed expert trajectories can often provide valuable guidance, e.g., plans and key actions, that can significantly improve agent exploration efficiency and acquisition of critical skills. Motivated by these observations, we propose Exploring Expert Failures (EEF), which identifies beneficial actions from failed expert trajectories and integrates them into the training dataset. Potentially harmful actions are meticulously excluded to prevent contamination of the model learning process. By leveraging the beneficial actions in expert failures, EEF successfully solves some previously unsolvable subtasks and improves agent tuning performance. Remarkably, our approach achieved a 62\% win rate in WebShop, outperforming RFT (53. 6\%) and GPT-4 (35. 6\%), and to the best of our knowledge, setting a new state-of-the-art as the first method to surpass a score of 0.81 in WebShop and exceed 81 in SciWorld.

Related papers

• ATLaS: Agent Tuning via Learning Critical Steps [39.279048406057264]
  Large Language Model (LLM) agents have demonstrated remarkable generalization capabilities across multi-domain tasks. Existing agent tuning approaches typically employ supervised finetuning on entire expert trajectories. We propose ATLaS that identifies the critical steps in expert trajectories and finetunes LLMs solely on these steps with reduced costs.
  arXiv  Detail & Related papers  (2025-03-04T02:14:55Z)
• From Novice to Expert: LLM Agent Policy Optimization via Step-wise Reinforcement Learning [62.54484062185869]
  We introduce StepAgent, which utilizes step-wise reward to optimize the agent's reinforcement learning process.<n>We propose implicit-reward and inverse reinforcement learning techniques to facilitate agent reflection and policy adjustment.
  arXiv  Detail & Related papers  (2024-11-06T10:35:11Z)
• Watch Every Step! LLM Agent Learning via Iterative Step-Level Process Refinement [50.481380478458945]
  Iterative step-level Process Refinement (IPR) framework provides detailed step-by-step guidance to enhance agent training. Our experiments on three complex agent tasks demonstrate that our framework outperforms a variety of strong baselines.
  arXiv  Detail & Related papers  (2024-06-17T03:29:13Z)
• ReAct Meets ActRe: When Language Agents Enjoy Training Data Autonomy [47.42940885853956]
  A$3$T is a framework that enables the Autonomous. of Agent Trajectories in the style of ReAct. In AlfWorld, the agent trained with A$3$T obtains a 1-shot success rate of 96%, and 100% success with 4 iterative rounds.
  arXiv  Detail & Related papers  (2024-03-21T17:43:44Z)
• Trial and Error: Exploration-Based Trajectory Optimization for LLM Agents [49.85633804913796]
  We present an exploration-based trajectory optimization approach, referred to as ETO. This learning method is designed to enhance the performance of open LLM agents. Our experiments on three complex tasks demonstrate that ETO consistently surpasses baseline performance by a large margin.
  arXiv  Detail & Related papers  (2024-03-04T21:50:29Z)
• Learning From Failure: Integrating Negative Examples when Fine-tuning Large Language Models as Agents [41.14201835950814]
  Large language models (LLMs) have achieved success in acting as agents, which interact with environments through tools such as search engines. Previous work has first collected interaction trajectories between LLMs and environments, using only trajectories that successfully finished the task to fine-tune smaller models. We argue that unsuccessful trajectories offer valuable insights, and LLMs can learn from these trajectories through appropriate quality control and fine-tuning strategies.
  arXiv  Detail & Related papers  (2024-02-18T17:10:07Z)
• Augmenting Unsupervised Reinforcement Learning with Self-Reference [63.68018737038331]
  Humans possess the ability to draw on past experiences explicitly when learning new tasks. We propose the Self-Reference (SR) approach, an add-on module explicitly designed to leverage historical information. Our approach achieves state-of-the-art results in terms of Interquartile Mean (IQM) performance and Optimality Gap reduction on the Unsupervised Reinforcement Learning Benchmark.
  arXiv  Detail & Related papers  (2023-11-16T09:07:34Z)
• Learning to Optimize for Reinforcement Learning [58.01132862590378]
  Reinforcement learning (RL) is essentially different from supervised learning, and in practice, these learneds do not work well even in simple RL tasks. Agent-gradient distribution is non-independent and identically distributed, leading to inefficient meta-training. We show that, although only trained in toy tasks, our learned can generalize unseen complex tasks in Brax.
  arXiv  Detail & Related papers  (2023-02-03T00:11:02Z)
• Learning from Guided Play: Improving Exploration for Adversarial Imitation Learning with Simple Auxiliary Tasks [8.320969283401233]
  We show that the standard, naive approach to exploration can manifest as a suboptimal local maximum. We present Learning from Guided Play (LfGP), a framework in which we leverage expert demonstrations of multiple exploratory, auxiliary tasks.
  arXiv  Detail & Related papers  (2022-12-30T20:38:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.