Overconfidence

2 items

RESEARCHarXiv CS.CL·4/14/2026

Self-Calibrating Language Models via Test-Time Discriminative Distillation

Large language models are often overconfident, expressing high certainty even when incorrect. This paper introduces SECL, a test-time training pipeline that exploits a self-supervised signal to improve calibration without requiring labeled data or human supervision.

Calibration self-supervision Overconfidence large language models

RESEARCHarXiv CS.AI·14d ago

Confidence Calibration in Large Language Models

This study investigates confidence calibration in Large Language Models (LLMs) across diverse tasks, finding that current LLMs are overconfident on difficult tests and underconfident on easy ones. The researchers developed LifeEval, a new test to evaluate model calibration across varying levels of difficulty.

Confidence Calibration Overconfidence machine learning large language models