From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning

arXiv cs.CL Papers
reinforcement-learning llm multi-agent-reasoning environment-design training-pipeline benchmark

Summary

This paper proposes the LLM-as-Environment-Engineer framework, where a policy model analyzes failures to automatically redesign the training environment for reinforcement learning, and introduces MAPF-FrozenLake as a controllable testbed. The framework, using Qwen3-4B, outperforms larger models like GPT and Gemini, showing that policy learning improves the model's ability to diagnose weaknesses.

arXiv:2606.17682v1 Announce Type: new Abstract: Reinforcement learning pipelines for Large Language Model (LLM) training often rely on manually redesigned environments between stages, requiring practitioners to heuristically infer which configuration will best improve the current policy. To automate this process, we propose the LLM-as-Environment-Engineer framework in which the current policy model analyzes failure trajectories together with contextual information and proposes modifications to the next-stage training environment configuration. We also introduce MAPF-FrozenLake, a controllable testbed whose generator exposes multi-dimensional environment configurations, making it suitable for studying and benchmarking environment redesign. On this testbed, we condition the environment engineer on structured summaries of policy behavior, failure cases, and environment statistics, from which it produces the configuration for the next training stage. With Qwen3-4B as the backbone, our framework achieves the strongest aggregate performance on our benchmarks, outperforming larger proprietary LLMs (e.g., GPT, Gemini) and fixed-environment training baselines. We further analyze which forms of context are most effective, finding that successful environment updates rely on failure evidence and preserve configurations that already work. Interestingly, the current RL checkpoint serves as a better environment engineer than the original base model, suggesting that policy learning improves the model's ability to diagnose its remaining weaknesses.
Original Article
View Cached Full Text

Cached at: 06/17/26, 05:41 AM 

# From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning
Source: [https://arxiv.org/abs/2606.17682](https://arxiv.org/abs/2606.17682)
[View PDF](https://arxiv.org/pdf/2606.17682)

> Abstract:Reinforcement learning pipelines for Large Language Model \(LLM\) training often rely on manually redesigned environments between stages, requiring practitioners to heuristically infer which configuration will best improve the current policy\. To automate this process, we propose the LLM\-as\-Environment\-Engineer framework in which the current policy model analyzes failure trajectories together with contextual information and proposes modifications to the next\-stage training environment configuration\. We also introduce MAPF\-FrozenLake, a controllable testbed whose generator exposes multi\-dimensional environment configurations, making it suitable for studying and benchmarking environment redesign\. On this testbed, we condition the environment engineer on structured summaries of policy behavior, failure cases, and environment statistics, from which it produces the configuration for the next training stage\. With Qwen3\-4B as the backbone, our framework achieves the strongest aggregate performance on our benchmarks, outperforming larger proprietary LLMs \(e\.g\., GPT, Gemini\) and fixed\-environment training baselines\. We further analyze which forms of context are most effective, finding that successful environment updates rely on failure evidence and preserve configurations that already work\. Interestingly, the current RL checkpoint serves as a better environment engineer than the original base model, suggesting that policy learning improves the model's ability to diagnose its remaining weaknesses\.

## Submission history

From: Chao Chen \[[view email](https://arxiv.org/show-email/2c74b96f/2606.17682)\] **\[v1\]**Tue, 16 Jun 2026 08:48:58 UTC \(4,921 KB\)

Similar Articles