# TopoCurate:Modeling Interaction Topology for Tool-Use Agent Training

arXiv: https://arxiv.org/abs/2603.01714 · HTML: https://mlsanity.com/paper/2603.01714/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 40 of its 47 references that are in the map.
- First posted on arXiv 2 Mar 2026.

## Builds on (most cited first, 30 of 40)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains](https://mlsanity.com/paper/2406.12045/) (2024)
- [DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models](https://mlsanity.com/paper/2512.02556/) (2025)
- [τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://mlsanity.com/paper/2506.07982/) (2025)
- [Kimi K2: Open Agentic Intelligence](https://mlsanity.com/paper/2507.20534/) (2025)
- [GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models](https://mlsanity.com/paper/2508.06471/) (2025)
- [APIGen-MT: Agentic Pipeline for Multi-Turn Data Generation via Simulated Agent-Human Interplay](https://mlsanity.com/paper/2504.03601/) (2025)
- [Where LLM Agents Fail and How They can Learn From Failures](https://mlsanity.com/paper/2509.25370/) (2025)
- [Jina Embeddings 2: 8192-Token General-Purpose Text Embeddings for Long Documents](https://mlsanity.com/paper/2310.19923/) (2023)
- [Agent Learning via Early Experience](https://mlsanity.com/paper/2510.08558/) (2025)
- [MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scaling](https://mlsanity.com/paper/2511.11793/) (2025)
- [Tree Search for LLM Agent Reinforcement Learning](https://mlsanity.com/paper/2509.21240/) (2025)
- [ACEBench: Who Wins the Match Point in Tool Usage?](https://mlsanity.com/paper/2501.12851/) (2025)
- [TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling](https://mlsanity.com/paper/2508.17445/) (2025)
- [Simulating Environments with Reasoning Models for Agent Training](https://mlsanity.com/paper/2511.01824/) (2025)
- [Towards General Agentic Intelligence via Environment Scaling](https://mlsanity.com/paper/2509.13311/) (2025)
- [Partially Observable Markov Decision Processes (POMDPs) and Robotics](https://mlsanity.com/paper/2107.07599/) (2021)
- [LongCat-Flash-Thinking-2601 Technical Report](https://mlsanity.com/paper/2601.16725/) (2026)
- [MUA-RL: Multi-turn User-interacting Agent Reinforcement Learning for agentic tool use](https://mlsanity.com/paper/2508.18669/) (2025)
- [ToolMind Technical Report: A Large-Scale, Reasoning-Enhanced Tool-Use Dataset](https://mlsanity.com/paper/2511.15718/) (2025)
- [Exploiting Tree Structure for Credit Assignment in RL Training of LLMs](https://mlsanity.com/paper/2509.18314/) (2025)
- [Synthesizing Post-Training Data for LLMs through Multi-Agent Simulation](https://mlsanity.com/paper/2410.14251/) (2024)
- [Procedural Environment Generation for Tool-Use Agents](https://mlsanity.com/paper/2506.11045/) (2025)
- [PACEvolve: Enabling Progress-Aware Consistent Evolution](https://mlsanity.com/paper/2601.10657/) (2026)
- [AutoForge: Automated Environment Synthesis for Agentic Reinforcement Learning](https://mlsanity.com/paper/2512.22857/) (2025)
- [ToolPRMBench: Evaluating and Advancing Process Reward Models for Tool-using Agents](https://mlsanity.com/paper/2601.12294/) (2026)
- [Pass@k Metric for RLVR: A Diagnostic Tool of Exploration, But Not an Objective](https://mlsanity.com/paper/2511.16231/) (2025)
- [WebOperator: Action-Aware Tree Search for Autonomous Agents in Web Environment](https://mlsanity.com/paper/2512.12692/) (2025)
- [Robust Tool Use via Fission-GRPO: Learning to Recover from Execution Errors](https://mlsanity.com/paper/2601.15625/) (2026)
- [Exploring Expert Failures Improves LLM Agent Tuning](https://mlsanity.com/paper/2504.13145/) (2025)

## Built on by (most cited first, 4 of 4)

- [Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application](https://mlsanity.com/paper/2606.12191/) (2026)
- [Hybrid Inspection and Task-Based Access Control in Zero-Trust Agentic AI](https://mlsanity.com/paper/2605.02682/) (2026)
- [Behavior Quotient Learning for Low-Rank Adaptation of LLM Agents](https://mlsanity.com/paper/2609.12896/) (2026)
- [Rollout-Time Intervention: Unlocking Agentic Reinforcement Learning Beyond Policy Limits](https://mlsanity.com/paper/2610.13438/) (2026)

## Abstract (arXiv)

Training tool-use agents typically relies on outcome-based filtering: Supervised Fine-Tuning (SFT) on successful trajectories and Reinforcement Learning (RL) on pass-rate-selected tasks. However, this paradigm ignores interaction dynamics: successful trajectories may lack error recovery or exhibit redundancy, while pass rates fail to distinguish structurally informative tasks from trivial ones. We propose TopoCurate, an interaction-aware framework that projects multi-trial rollouts from the same task into a unified semantic quotient topology. By merging equivalent action-observation states, this projection transforms scattered linear trajectories into a structured manifold that explicitly captures how tool invocations and environmental responses drive the divergence between effective strategies and failure modes. Leveraging this representation, we introduce a dual-selection mechanism: for SFT, we prioritize trajectories demonstrating reflective recovery, semantic efficiency, and strategic diversity to mitigate covariate shift and mode collapse; for RL, we select tasks with high error branch ratios and strategic heterogeneity, maximizing gradient Signal-to-Noise Ratio to address vanishing signals in sparse-reward settings. Evaluations on BFCLv3 and Tau2 Bench show that TopoCurate achieves consistent gains of 4.2% (SFT) and 6.9% (RL) over state-of-the-art baselines. We will release the code and data soon for further investigations.