AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Knowledge Distillation for Large Language Models through Residual Learning

Thinh On, Hengzhi Pei, Leonard Lausen, George Karypis

ICLR 2026 · Label-efficient and robust learning › Knowledge distillation · approach: Rethinking distillation objectives

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 scaling factor beta in residual computation (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 16.25 -> 20.01Improved
add teacher-error accuracy mask (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 19.59 -> 20.01Improved
add pretraining of teacher projector (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 17.39 -> 20.01Improved
freeze the pretrained teacher projector during distillation (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 19.07 -> 20.01Improved
add expert knowledge fusion instead of sparse 2/8 activation (ablation)MoE teacher · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 19.66 -> 20.01Improved
add expert knowledge fusion instead of dense 8/8 activation (ablation)MoE teacher · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 19.68 -> 20.01Improved
add expert knowledge fusion instead of average pooling (ablation)MoE teacher · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 18.43 -> 20.01Improved
increase alignment dimension dA from 64 to 768 (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 20.01 -> 18.02Hurt
add residual learning to DSKD (ablation)DSKD · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 18.40 -> 19.60Improved
use residual learning instead of DSKD with reverse KL (ablation)DSKD · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 15.98 -> 20.01Improved
lower lambda from 0.5 to 0.1 (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 20.01 -> 16.39Hurt
raise lambda from 0.5 to 0.9 (ablation)residual learning · Instruction following avg, Mixtral-8x7B -> GPT2-120M (cross-tokenizer)Avg Rouge-L: 20.01 -> 18.81Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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