Early-Signal Detector

method · first seen 4 Oct 2026

supervised fine-tuning

Also: SFT

Training a pretrained model further on labelled input-output examples to adapt its behaviour to a target task.

Mentions per week

01619 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Rank 107 on 16 Nov 2026 · 26 mentions in 28 days · 26 independent groups · papers → code · first seen 6 weeks ago · acceleration +0.33

Earliest sources

Latest sources

Mentioned together with

reinforcement learning with verifiable rewardscatastrophic forgettinggroup relative policy optimizationtool callingdirect preference optimizationlow-rank adaptationon-policy self-distillationHealthBenchreinforcement learning from task feedbackcheckpoint weight-update auditing