TreeFlex: Agents Navigating Their Own Search Trees
We introduce TreeFlex, a framework in which the agent's own policy decides where to search.
ProofPaper ↗
Key points
- A ReAct-style agent accumulates observations, intermediate results, and failed attempts in a single context.
- The agent records each intermediate state, with a short summary, as a node in a search tree maintained by the harness.
- We train navigation, task actions, checkpointing, and proposal review jointly with a tree-aware adaptation of GRPO, using task-level outcome rewards augmented by process rewards.
- On SWE-bench Verified, with all three systems trained with RL from the same 9B model, TreeFlex resolves 54.9% of tasks, versus 43.3% for FoldAgent and 45.2% for its step-wise variant.
Sources (1)
- [1]TreeFlex: Agents Navigating Their Own Search TreesarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 14, 03:16 PM
We introduce TreeFlex, a framework in which the agent's own policy decides where to search.
A ReAct-style agent accumulates observations, intermediate results, and failed attempts in a single context.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 14, 2026Beyond the Trial-and-Error Loop: Hybrid Projection and Automated Tuning for Distributed Training
- Oct 14, 2026More Methods, Better Agents? A Reality Check on Self-Distillation in Agentic Reinforcement Learning
- Oct 13, 2026Knowledge Graphs as Files: An Agent Harness Where Tool Creation and Tool Use Co-Evolve
- Oct 13, 2026How Low Can You Go: Fine-Tuning LLMs with Bayesian Optimization (LlMBO)
- Oct 13, 2026Copying Is Not Grounding: Training Faithful Long-Form RAG with Authored-Sentence Credit
- Oct 13, 2026Benchmarking and Training Reward Models for Cross-Source Evidence Grounding in Search Agents