AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

ToolACE-R: Model-aware Iterative Training and Adaptive Refinement for Tool learning

Xingshan Zeng, Weiwen Liu, Xu Huang, Zezhong Wang, Lingzhi Wang, Liangyou Li, Yasheng Wang, Lifeng Shang, Xin Jiang, Ruiming Tang, Qun Liu

AAAI 2026 · Large language models › Instruction and fine-tuning · approach: Efficient LLM fine-tuning systems · 16 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
remove model-aware data selection (ablation)ToolACE-R · BFCL Overallaccuracy: 86.33 -> 85.04Improved
remove self-refinement data (ablation)ToolACE-R · BFCL Overallaccuracy: 85.04 -> 84.42Improved
add adaptive self-refinement at inference (ablation)ToolACE-R · BFCL Overallaccuracy: 86.33 -> 86.83Improved
add adaptive self-refinement at inferenceToolACE-R · ACEBenchaccuracy: 83.88 -> 84.00No clear effect
add adaptive self-refinement at inferenceToolACE-R · API-Bankaccuracy: 75.89 -> 81.22Improved
include extremely easy samples (alpha=-1) (ablation)ToolACE-R · BFCL Overallaccuracy: 86.33 -> 86.00Hurt
keep only difficult samples (alpha=0.1, beta=2) (ablation)ToolACE-R · BFCL Overallaccuracy: 86.33 -> 85.31Hurt
use k=8 attempts instead of k=1 for difficulty (ablation)ToolACE-R · BFCL Overallaccuracy: 85.96 -> 86.33Improved
select data with the model being trained instead of Qwen2.5-3B (ablation)Llama3.1-8B-Inst · BFCL Overallaccuracy: 85.50 -> 86.33Improved
use adaptive stopping instead of vanilla self-refine (ablation)ToolACE-R · BFCLaccuracyImproved
apply ToolACE-R training to Mistral-7B-Inst-v0.2 vs original-data SFTMistral-7B-Instruct-v0.2 · BFCL Overallaccuracy: 77.12 -> 84.00Improved
apply ToolACE-R training to Qwen2.5-7B-Inst vs original-data SFTQwen2.5-7B-Instruct · BFCL Overallaccuracy: 83.34 -> 85.21Improved

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