LLM-generated training datalabel noisepublished Nov 14, 2026 · arXiv:2611.16498
How many LLM-labelled rows does a decision encoder need, and how wrong can the labels be?
confidence calibrationtemperature scalingpublished Nov 14, 2026 · arXiv:2611.16499
Recalibrating a Typed-Decision Model in an Afternoon: A Hands-On Recipe for Laya's Confidence Scores
agent harnessesSystem-1 decision modelspublished Oct 5, 2026 · arXiv:2610.02267
Fast Models, Slow Evidence: A Paired and Self-Audited Evaluation of System-1 Decision Models for LLM Agent Harnesses
decision modelsvalue of informationpublished Sep 28, 2026 · arXiv:2609.30706
LAVOIR: Teaching a Single-Pass Decision Encoder When and What to Ask with Amortized Value of Information