AI Paper Atlas

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

From Parameters to Data: A Task-Parameter-Guided Fine-Tuning Pipeline for Efficient LLM Alignment

Hao Chen, Qi Zhang, Liyao Li, Zhanming Shen, Wentao Ye, Lirong Gao, NINGTAO WANG, Xing Fu, Xiaoyu Shen, Junbo Zhao

ICML 2026 · Large language models › Instruction and fine-tuning · approach: Efficient LLM fine-tuning systems · 0 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
add P2D parameter-guided data selection instead of Data Whisperersparse head adaptation · Average over 3 datasets x 3 modelsaverage score: 49.38 -> 51.87Improved
add P2D sparse head adaptation (10% heads) instead of LoRA on full datafull data fine-tuning · Average over 3 datasets x 3 modelsaverage score: 54.25 -> 56.22Improved
update 10% of heads chosen by P2D instead of random heads (ablation)full data fine-tuning · Average over 3 datasets x 3 modelsaverage score: 46.64 -> 56.22Improved
add P2D sparse head adaptation to P2D data selection versus full parameters (ablation)P2D data selection · Average over 3 datasets x 3 modelsaverage score: 50.58 -> 51.87Improved
reduce data ratio from 100% to 10% (GSM8K, Qwen-2.5-7B) (ablation)P2D · GSM8KEM: 83.03 -> 82.56No clear effect
reduce data ratio from 100% to 10% (BioInstruct, Qwen-2.5-7B) (ablation)P2D · BioInstructROUGE-L: 41.92 -> 32.78Hurt
reduce data ratio from 100% to 10% (DialogSum, Llama-3-8B) (ablation)P2D · DialogSumROUGE-L: 48.15 -> 42.45Hurt
reduce head ratio from 100% to 10% (GSM8K, Qwen-2.5-7B) (ablation)P2D · GSM8KEM: 80.05 -> 82.56Improved
reduce head ratio from 100% to 10% (BioInstruct, Qwen-3-8B) (ablation)P2D · BioInstructROUGE-L: 33.45 -> 32.89Hurt
scale model from 1.5B to 32B with P2D sparse adaptation versus Full SFT (GSM8K)P2D-sparse head adaptation · GSM8K (full data)EM: 63.45 -> 65.23 at 1.5B; 88.94 -> 90.23 at 32BImproved
add P2D sparse adaptation versus Full SFT on general benchmarks after UltraChatQwen-2.5-7B · MMLUaccuracy: 56.34 -> 67.45Improved
use W1 head scoring instead of cosine similarity (ablation)P2D · GSM8KEM: 63.28 -> 78.82Improved

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