AI Paper Atlas

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

Explain in Your Own Words: Improving Reasoning via Token-Selective Dual Knowledge Distillation

Minsang Kim, Seung Jun Baek

ICLR 2026 · Label-efficient and robust learning › Knowledge distillation · approach: Distillation with causal and adaptive guidance · 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
add token-selective sigmoid gating in direct distillation (ablation)on-policy GKD · MMLU-STEM / IFEval (Qwen2.5 1.5B)accuracy: IFEval 52.3 -> 54.2; MMLU-STEM 47.7 -> 48.1Improved
add indirect preference-ranking distillation (ablation)TSD-KD with gating · MATH (Qwen2.5 14B->1.5B)accuracy: 18.2 -> 20.9Improved
add selective entropy minimization (ablation)TSD-KD with gating · MATH (Qwen2.5 14B->1.5B)accuracy: 18.2 -> 22.3Improved
combine all three components (ablation)on-policy GKD baseline · MATH (Qwen2.5 14B->1.5B)accuracy: 18.1 -> 26.1Improved
add entropy minimization on top of gating, indirect off (ablation)TSD-KD · GSM-Plus (Qwen2.5 14B->1.5B)accuracy: 40.3 -> 39.8No clear effect
use on-policy rather than off-policy data (ablation)TSD-KD · Qwen2.5 14B->1.5B, six tasksaccuracy: GSM8K 56.0 -> 60.1; MATH 24.9 -> 26.1Improved
distill all tokens in opener (c=100%) instead of c=10% (ablation)TSD-KD indirect distillation · average over benchmarksaccuracyHurt
increase top-k in indirect distillation from 10 to 20 (ablation)TSD-KD · average over benchmarksaccuracy: -> 44.5Hurt
apply entropy minimization to all tokens (s=100%) instead of top 10% (ablation)TSD-KD · average over benchmarksaccuracy: 45.7 -> 41.8Hurt
raise JSD beta from 0 (forward KL) to 0.9 (ablation)TSD-KD direct distillation · average over benchmarksaccuracy: 43.4 -> 45.7Improved
use adaptive c instead of fixed c=10% (ablation)TSD-KD · MATH (Qwen2.5 14B->1.5B)accuracy: MATH 26.1 -> 26.5; MBPP 42.1 -> 39.9Mixed
apply TSD-KD with LoRA instead of full finetuningTSD-KD · PEFT (Qwen2.5 1.5B)accuracy: MMLU-STEM 49.7 vs GKD 47.6Improved

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