AI Paper Atlas

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

Grokking in LLM Pretraining? Monitor Memorization-to-Generalization without Test

Ziyue Li, Chenrui Fan, Tianyi Zhou

ICLR 2026 · Large language models › Language, safety and interpretability · approach: LLM pruning and scaling data · 13 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
monitor pathway consistency instead of training loss as generalization indicatortest-accuracy correlation · MathPearson correlation: -0.0435 -> 0.9246Improved
monitor pairwise pathway distance instead of validation losstest-accuracy correlation · CommonsensePearson correlation: -0.5752 -> -0.9821Improved
monitor pathway consistency instead of training losstest-accuracy correlation · Domain-SpecificPearson correlation: 0.7944 -> 0.9917Improved
increase LoRA rank from 32 to 64 (ablation)pathway distance correlation · CommonsensePearson correlation: -0.9821 -> -0.9927No clear effect
increase LoRA rank from 32 to 64 (ablation)pathway consistency correlation · CommonsensePearson correlation: 0.9804 -> 0.9559No clear effect
vary routing-weight threshold from 0.5 to 0.8 (ablation)pathway edit distance trend · Domain-specific QAedit distance trendNo clear effect
vary loss threshold epsilon from 1.5 to 2.0 (ablation)pathway edit distance trend · Domain-specific QAedit distance trendNo clear effect
vary stability threshold delta between 0.03 and 0.05 (ablation)pathway edit distance trend · Domain-specific QAedit distance trendNo clear effect
router weights versus expert weights update speedconvergence speed · OLMoE routing dynamicsconvergence speed: 0.151 vs 0.063Mixed

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