Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback

Tian et al. (Sun,) studied this question.

www.synapsesocial.com/papers/69dd247f03ecda3005e52cf6 — DOI: https://doi.org/10.18653/v1/2023.emnlp-main.330

Also consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

1. On Calibration of Modern Neural Networks· 2017 · 1,727 citations
2. Language Models (Mostly) Know What They Know· 2022 · 161 citations
3. Training language models to follow instructions with human feedback· 2022 · 4,276 citations
4. Selective Classification for Deep Neural Networks· 2017 · 174 citations
5. Overcoming the Inevitable Anchoring Effect: Considering the Opposite Compensates for Selective Accessibility· 2000 · 556 citations