AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

Log Probability Tracking of LLM APIs

Timothee Chauvin, Erwan Le Merrer, Francois Taiani, Gilles Tredan

ICLR 2026 · Large language models › Instruction and fine-tuning · approach: Fine-tuning generalization and delta compression · 5 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
use a longer prompt (33 tokens instead of 1.5 tokens) (ablation)Logprob tracking · TinyChangeROC AUC: about 1% AUC difference, baseline around 91% AUCNo clear effect
use logprob tracking instead of METchange detection · TinyChangeROC AUC: 0.670 -> 0.915Improved
use logprob tracking instead of MMLU-ALGchange detection · TinyChangeROC AUC: 0.878 -> 0.915Improved
use logprob tracking instead of MET (cost)change detection · TinyChangecost per year: $146 -> $0.14Improved
use logprob tracking instead of MET (sensitivity, weight pruning)change detection · TinyChange weight pruningthreshold pruned fraction for AUC 0.9: 2^-1 -> 2^-10 (or below)Improved
use logprob tracking instead of MMLU-ALG (sensitivity, weight pruning)change detection · TinyChange weight pruningthreshold pruned fraction for AUC 0.9: 2^-4 -> 2^-10 (or below)Improved
detect LoRA fine-tuning instead of full fine-tuningLogprob tracking · TinyChangeROC AUC: L2 distance 1.3 x 10^-5 -> 4.3 x 10^-9 (1 step)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