AI Paper Atlas

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

OPIC: Enhancing Language Model Merging via Optimizing In-Context Capability

Jie He, Weidong Bao, Chao Chen, Zhengyi Zhong, Shuai Zhang, Ji Wang

ICML 2026 · Large language models › Language, safety and interpretability · approach: LLM pruning and scaling data

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 inter-layer refinement matrix A (ablation)OPIC · 4-task merge, Qwen2.5-1.5Bmean normalized score: 77.80% -> 80.70%Improved
add task scaling vector B (ablation)OPIC · 4-task merge, Qwen2.5-1.5Bmean normalized score: 78.30% -> 80.70%Improved
add both A and B (hierarchical refinement) beyond global scalar C (ablation)OPIC · 4-task merge, Qwen2.5-1.5Bmean normalized score: 76.40% -> 80.70%Improved
widen refinement search range from [0,1] to [0,1.2] and [0,1.5] (ablation)OPIC · 4-task merge, Qwen2.5-1.5Bnormalized score: max difference 0.36%No clear effect
use OPIC validation-free optimization instead of DOGEOPIC vs DOGE · 4-task merge, Qwen2.5-1.5Bmean normalized score: 69.62% -> 80.73%Improved
integrate OPIC into TIESTIES · 4-task merge, Qwen2.5-1.5Bmean normalized score: 77.63% -> 78.80%No clear effect
integrate OPIC into DAREDARE · 4-task merge, Qwen2.5-1.5Bmean normalized score: 78.32% -> 76.10%No clear effect
integrate OPIC into TIESTIES · ANLI, Qwen2.5-1.5Bnormalized score: ANLI 80.0% -> 100.1%; COQA 58.9% -> 49.0%Mixed
TIES and DARE pruning-based merging versus simple Task ArithmeticTIES/DARE · 4-task merge, Qwen2.5-1.5Bmean normalized score: 79.56% -> 77.63% (TIES), 78.32% (DARE)Hurt
OPIC versus TIES/DARE on out-of-domain reasoningOPIC · ARC, Qwen2.5-1.5Baccuracy: 0.761 (TIES) vs 0.653Hurt
OPIC on MMLU versus Task ArithmeticOPIC · MMLU, Qwen2.5-1.5Baccuracy: 0.332 -> 0.544Improved
OPIC evolutionary search versus grid search computeOPIC · merge costGPU hours: ~0.2 -> ~0.54Hurt

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