AI Paper Atlas

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

Importance-Aware Data Selection for Efficient LLM Instruction Tuning

Tingyu Jiang, Shen Li, Yiyao Song, Lan Zhang, Hualei Zhu, Yuan Zhao, Xiaohang Xu, Kenjiro Taura, Hao Henry Wang

AAAI 2026 · Large language models › Instruction and fine-tuning · approach: Instruction data selection · 9 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
select top-MIWV samples instead of training on full dataset (1% of Alpaca)LLaMA2-7B instruction tuning · AlpacaEvalwin rate: 27.75 -> 39.50Improved
select top-MIWV samples instead of full data (1% of Alpaca)LLaMA2-13B instruction tuning · AlpacaEvalwin rate: 35.00 -> 41.30Improved
select 1% of WizardLM data by MIWV instead of full dataLLaMA2-13B · Open LLM Leaderboard averageaverage score: 60.23 -> 60.30No clear effect
select data with random sampling instead of MIWV (ablation)LLaMA-7B on Alpaca · pairwise win rate vs official Alpacawin rateHurt
select data with high prompt loss instead of high MIWV (ablation)LLaMA-7B on Alpaca · pairwise win rate vs official Alpacawin rateHurt
select low MIWV instead of high MIWV samples (ablation)LLaMA-7B on Alpaca · pairwise win rate vs official Alpacawin rateHurt
swap embedding model for retrieval (Bge-en-large vs Gte-base-en-v1.5) (ablation)MIWV on LLaMA-7B Alpaca · pairwise win ratewin rateNo clear effect
add ICL one-shot retrieval to IFD Score (ablation)IFD Score on LLaMA-7B Alpaca 1% · five test sets win ratewin rate: 0.939 -> 1.017Improved
use MIWV scoring instead of ICL + IFD Score (ablation)Alpaca 1% LLaMA-7B · five test sets win ratewin rate: 1.017 -> 1.140Improved
MIWV selection vs other selection methods at 1% scaleLLaMA2-7B Alpaca · WizardLM test setwin rate vs full data: 1.119 vs 1.092 (RECOST)Improved
increase fraction of selected data toward the full set (ablation)MIWV subsets · pairwise win ratewin rateHurt
apply MIWV to Qwen2.5-7B/14BQwen2.5 instruction tuning · LIMA test setwin rateImproved

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