# Balancing Classification and Calibration Performance in Decision-Making LLMs via Calibration Aware Reinforcement Learning

arXiv: https://arxiv.org/abs/2601.13284 · HTML: https://mlsanity.com/paper/2601.13284/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 41 references that are in the map.
- First posted on arXiv 19 Jan 2026.

## Builds on (most cited first, 24 of 24)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering](https://mlsanity.com/paper/1809.02789/) (2018)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://mlsanity.com/paper/2312.06674/) (2023)
- [Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback](https://mlsanity.com/paper/2305.14975/) (2023)
- [Teaching Models to Express Their Uncertainty in Words](https://mlsanity.com/paper/2205.14334/) (2022)
- [WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs](https://mlsanity.com/paper/2406.18495/) (2024)
- [Open-Reasoner-Zero: An Open Source Approach to Scaling Up Reinforcement Learning on the Base Model](https://mlsanity.com/paper/2503.24290/) (2025)
- [Measuring Calibration in Deep Learning](https://mlsanity.com/paper/1904.01685/) (2019)
- [Verified Uncertainty Calibration](https://mlsanity.com/paper/1909.10155/) (2019)
- [Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters](https://mlsanity.com/paper/2212.10001/) (2022)
- [Kernel Language Entropy: Fine-grained Uncertainty Quantification for LLMs from Semantic Similarities](https://mlsanity.com/paper/2405.20003/) (2024)
- [Beyond Binary Rewards: Training LMs to Reason About Their Uncertainty](https://mlsanity.com/paper/2507.16806/) (2025)
- [Taming Overconfidence in LLMs: Reward Calibration in RLHF](https://mlsanity.com/paper/2410.09724/) (2024)
- [SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales](https://mlsanity.com/paper/2405.20974/) (2024)
- [Batch Calibration: Rethinking Calibration for In-Context Learning and Prompt Engineering](https://mlsanity.com/paper/2309.17249/) (2023)
- [Determinants of LLM-assisted Decision-Making](https://mlsanity.com/paper/2402.17385/) (2024)
- [MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMs](https://mlsanity.com/paper/2402.11756/) (2024)
- [Restoring Calibration for Aligned Large Language Models: A Calibration-Aware Fine-Tuning Approach](https://mlsanity.com/paper/2505.01997/) (2025)
- [Do Not Design, Learn: A Trainable Scoring Function for Uncertainty Estimation in Generative LLMs](https://mlsanity.com/paper/2406.11278/) (2024)
- [On Calibration of LLM-based Guard Models for Reliable Content Moderation](https://mlsanity.com/paper/2410.10414/) (2024)
- [Calibrated Language Models and How to Find Them with Label Smoothing](https://mlsanity.com/paper/2508.00264/) (2025)
- [Uncertainty as Feature Gaps: Epistemic Uncertainty Quantification of LLMs in Contextual Question-Answering](https://mlsanity.com/paper/2510.02671/) (2025)

## Built on by (most cited first, 3 of 3)

- [OmniMed-Jev: Calibrating LVLM Confidence for Trustworthy Medical Multimodal Decisions via System One](https://mlsanity.com/paper/2610.00381/) (2026)
- [ufakzeka-karar: An Open Turkish Typed-Decision Model with Order-Invariant Option Scoring](https://mlsanity.com/paper/2610.06744/) (2026)
- [Can an Agent Learn to Predict Its Own Failure? Outcome-Grounded Confidence Calibration in LLM Agents](https://mlsanity.com/paper/2610.18614/) (2026)

## Abstract (arXiv)

Large language models (LLMs) are increasingly deployed in decision-making tasks, where not only accuracy but also reliable confidence estimates are essential. Well-calibrated confidence enables downstream systems to decide when to trust a model and when to defer to fallback mechanisms. In this work, we conduct a systematic study of calibration in two widely used fine-tuning paradigms: supervised fine-tuning (SFT) and reinforcement learning with verifiable rewards (RLVR). We show that while RLVR improves task performance, it produces extremely overconfident models, whereas SFT yields substantially better calibration, even under distribution shift, though with smaller performance gains. Through targeted experiments, we diagnose RLVR's failure, showing that decision tokens act as extraction steps of the decision in reasoning traces and do not carry confidence information, which prevents reinforcement learning from surfacing calibrated alternatives. Based on this insight, we propose a calibration-aware reinforcement learning formulation that directly adjusts decision-token probabilities. Our method preserves RLVR's accuracy level while mitigating overconfidence, reducing ECE scores up to 9 points.