Hoppa till innehÄllet
Global sound
Daily Paper Cast

Daily Paper Cast

ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents

· 23 min · Avsnitt 2473

Avsnittsbeskrivning

đŸ€— Upvotes: 30 | cs.CL

Authors:

Guanting Dong, Zhichao Sun, Xinyu Yang, Rui Wang, Xin Deng, Jiajie Jin, Zhixiong Yang, Junhao Huang, Han Han, Mingyu Ma, Xiaotao Wang, Yuyang Hu, Xiaoshuai Song, Xiaoxi Li, Jiejun Tan, Zhecan Wang, Shihan Dou, Honglin Guo, Zhiheng Xi, Tao Gui, Tao Ji, Qi Zhang, Yutao Zhu, Ji-Rong Wen, Zhicheng Dou

Title:

ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents

Arxiv:

http://arxiv.org/abs/2611.05036v1

Abstract:

Large language models are deployed as long-horizon agents that compose diverse tools across MCP services into executable workflows. Existing MCP environment-scaling methods are data-centric: they first synthesize a database and scenario, then derive tools and tasks from them. As a result, tool interfaces may not match real ones, tool compositions are not grounded in execution, and tasks rarely span more than one environment. We propose ToolSpace, a tool-centric framework that reverses this order. ToolSpace mines data, state, and retry dependencies from adjacent calls in execution traces of real MCP services and open agent corpora, using deterministic rules rather than LLM judgment, and clusters the resulting weighted tool-composition graph into environments. Databases are synthesized only for state-mutating environments, while query and stateless tools replay recorded returns. ToolSpace then performs state-aware random walks with rollback and decompiles each chain into a natural-language task with assertion-based checks, keeping only samples that replay successfully from the initial state. Applied to AutomationBench and tau3-Bench Banking, the pipeline produces harder Pro versions on which 7 frontier models drop by up to 38.2 points. Across seven long-horizon tool-use benchmarks, ToolSpace raises Qwen3.6-35B-A3B from 23.0 to 41.0 average score (43.2 with RL) and GLM-5.2 from 44.6 to 56.6, on par with Claude Opus 5. Ablations show that execution-witnessed edges clearly outperform LLM-judged ones, and performance keeps improving with data and environment scale.

Webbplats: share.transistor.fm

Live, senaste timmen: 12 888 lÄtar pÄ 6 286 stationer i 132 lÀnder

UpptÀck vÀrlden genom radion

Fler sÀtt att upptÀcka (7)

Hört nÄgot bra? Spara det, eller öppna det direkt i Spotify.

★ Artist of the week The fastest riser on the radio this week
Your song here? From € 9 per week →
Radio ArasuncionFortsÀtt lyssna