AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

ToolWeaver: Weaving Collaborative Semantics for Scalable Tool Use in Large Language Models

Bowen Fang, Wen Ye, Yunyue Su, Jinghao Zhang, Qiang Liu, Yesheng Liu, Xin Sun, Shu Wu, Jiabing Yang, Baole Wei, Liang Wang

ICLR 2026 · Large language models › Language, safety and interpretability · approach: LLM pruning and scaling data · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace atomic one-token-per-tool IDs with collaborative-aware hierarchical codesgenerative tool retrieval (Llama-3-8B) · ToolBench retrieval I3 (multi-domain)NDCG@1: 81.00 -> 88.00Improved
replace atomic IDs with ToolWeaver codes for end-to-end tool useToolGen-style agent (Llama-3-8B) · ToolBench end-to-end I3SoPR: 36.34 -> 52.19Improved
replace atomic IDs with ToolWeaver codesToolGen-style agent (Llama-3-8B) · ToolBench end-to-end I2SoPR: 45.13 -> 44.03Hurt
replace atomic IDs with ToolWeaver codesToolGen-style agent (Llama-3-8B) · ToolBench end-to-end I1SoPR: 52.97 -> 53.17No clear effect
replace atomic IDs with ToolWeaver codesToolGen-style agent (Llama-3-8B) · ToolBench retrieval I3 (in-domain)NDCG@1: 87.00 -> 86.00No clear effect
add collaborative guidance (graph Laplacian regularization) to the tokenizer (ablation)ToolWeaver tokenization · ToolBench retrieval (I1-I3)NDCG@kImproved
add semantic initialization from a text encoder (ablation)ToolWeaver tokenization · ToolBench retrieval (I1-I3)NDCG@k: over 20 NDCG pointsImproved
increase collaborative weight lambda from 0.01 to 1 (ablation)ToolWeaver tokenization · ToolBench retrieval (avg NDCG)NDCG@kImproved
increase collaborative weight lambda from 1 to 10 (ablation)ToolWeaver tokenization · ToolBench retrieval (avg NDCG)NDCG@kHurt
use hierarchical or semantic naming codes instead of atomic tokens (ablation)ToolGen-style agent (Llama-3-8B) · ToolBench end-to-end I3SoPR: 36.34 -> 17.49 (hierarchical), 16.39 (semantic)Hurt
use linear atomic vocabulary expansion (ToolGen) rather than logarithmic code expansionLlama-3-8B · WikiText-2perplexity: 25.36 -> 104.54Hurt
add ToolWeaver codes to Llama-3-8BLlama-3-8B · CNN/DailyMail summarizationBERTScore F1: 85.35 -> 85.07No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy