AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

MoSE: Hierarchical Self-Distillation Enhances Early Layer Embeddings

Andrea Gurioli, Federico Pennino, Joao Monteiro, Maurizio Gabbrielli

AAAI 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Layer-wise pruning of LLMs

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 multi-layer self-distillation loss at fine-tuning (vs single-exit fine-tuning) (ablation)MoSE exit heads · CodeSearchNetRecall@1: up to +4.36% averaged Recall@1 at layer 9Improved
exit at layer 4 instead of layer 36MoSE · CodeSearchNet text-to-codeavg MRR: 6.4% absolute drop for about 90% fewer FLOPsHurt
use shallow layers for code-to-code retrievalMoSE · POJ104mAPImproved
use layer 18 for text-to-code retrievalMoSE · CodeSearchNetMRRImproved
replace NSP with In-Context Classification in pretraining (ablation)MoSE-30M · CoIR CSN-CCRNDCG@10: 6.2 -> 10.1Improved
replace NSP with In-Context Classification in pretraining (ablation)MoSE-30M · CoIR CoSQA, CT-DL, CT-ContestNDCG@10: 31.0 -> 32.0 on CT-DL; 12.0 -> 12.1 on CT-Contest; 0.4 on CoSQANo clear effect
go deeper than layer 18 on code clone detectionMoSE · BigCloneBenchF1: 94.2 (L18) vs 94.1 (L36)No clear effect
use very early exits (L4, L9) for clone detectionMoSE · BigCloneBenchprecision: 89.8 (L4), 90.4 (L9) vs 91.7 (L36)Hurt

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