[2003.07892] Calibration of Pre-trained Transformers

Abstract:Pre-trained Transformers are now ubiquitous in natural language processing, but despite their high end-task performance, little is known empirically about whether they are calibrated. Specifically, do these models' posterior probabilities provide an accurate empirical measure of how likely the model is to be correct on a given example? We focus on BERT and RoBERTa in this work, and analyze their calibration across three tasks: natural language inference, paraphrase detection, and commonsense reasoning. For each task, we consider in-domain as well as challenging out-of-domain settings, where models face more examples they should be uncertain about. We show that: (1) when used out-of-the-box, pre-trained models are calibrated in-domain, and compared to baselines, their calibration error out-of-domain can be as much as 3.5x lower; (2) temperature scaling is effective at further reducing calibration error in-domain, and using label smoothing to deliberately increase empirical uncertainty helps calibrate posteriors out-of-domain.

Abstract:Pre-trained Transformers are now ubiquitous in natural language processing, but despite their high end-task performance, little is known empirically about whether they are calibrated. Specifically, do these models' posterior probabilities provide an accurate empirical measure of how likely the model is to be correct on a given example? We focus on BERT and RoBERTa in this work, and analyze their calibration across three tasks: natural language inference, paraphrase detection, and commonsense reasoning. For each task, we consider in-domain as well as challenging out-of-domain settings,

related reading

• Transformer Circuits Threadtransformer-circuits.pub
• gpt-4.pdfcdn.openai.com
• Transformer Explainer: LLM Transformer Model Visually Explainedpoloclub.github.io
• [1706.04599] On Calibration of Modern Neural Networksarxiv.org
• [2603.02202] Frontier Models Can Take Actions at Low Probabilitiesarxiv.org
• Understanding Transformers... (beyond the Math) – kalomaze's kalomazing blogkalomaze.bearblog.dev
• Large Language Models Reading List | Sebastian Raschka, PhDsebastianraschka.com
• Getting Caught Up to Modern LLM Research | Samarth Goeldev.samarthgoel.com
• On Calibration of Modern Neural Networksproceedings.mlr.press
• Recent Advances in Language Model Fine-tuningruder.io
• [2506.19733] Breaking Barriers: Do Reinforcement Post Training Gains Transfer To Unseen Domains?arxiv.org
• Generalized Language Models | Lil'Loglilianweng.github.io