← Back to all news
arXiv Computation and Language By Davide Baldelli, Sruthi Kuriakose, Maryam Hashemzadeh, Amal Zouaq, Sarath Chandar

Probabilistic Calibration Is a Trainable Capability in Language Models

Read the original on arXiv Computation and Language →

The paper investigates whether language models can be fine‑tuned to produce outputs that better match intended probability distributions, even when the valid outputs cannot be enumerated in advance. Two calibration fine‑tuning methods—soft‑target and hard‑target—are evaluated across 12 models from four families, showing substantial improvements in structured‑sampling fidelity on unseen distribution families and parameter settings. The gains also transfer to broader stochastic generation tasks, expanding first‑token support and improving NoveltyBench scores, though some downstream capabilities such as arithmetic reasoning may be slightly affected.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computation and Language.

Related stories

arXiv AI
Oct 5

Test-time Calibration Learning for Large Language Model Reasoning

The paper introduces Test-Time Calibration Learning (TTCL), a label‑free framework that adapts both reasoning accuracy and verbalized confidence of large language models directly on unlabeled target‑task data. TTCL generates self‑supervision signals from multiple model responses, enabling calibration without ground‑truth labels and proving theoretically as a bounded surrogate for the ideal calibration objective. Experiments on mathematical reasoning and factual question answering show consistent improvements, with base models gaining an average 40.13% in accuracy and 70.80% in ECE reduction across eight benchmarks, and further gains under domain shift.

By Zizhuo Zhang, Xiong Peng, Jingwei Sun, Rong Yao, Shixiong Kai, Mingxuan Yuan, Bo Han
llmsnlpreinforcement-learningbenchmarks
More like this →
arXiv AI
Aug 28

Mutual Debiasing via Dual-Seed Comparison for Probabilistic Sampling in Large Language Models

The paper introduces Dual-Seed Comparison (DSC), a protocol that uses two independent LLM-generated seeds to reduce systematic bias in probabilistic sampling. DSC constructs a bit sequence from the character-level ordinal values of the seeds, normalizes it into a pseudo-uniform variate, and maps it to the target distribution via the inverse cumulative distribution function. Empirical results show DSC outperforms existing methods in 96% of evaluated settings and enhances distributional control in tasks like MCQ generation and attribute-constrained text-to-image prompting.

By Zihao Guo, Hongtao Lv, Chaoli Zhang, Laiguo Yin, Lei Liu, Yonghui Xu, Lizhen Cui
llmsdiffusionsafety
More like this →
arXiv AI
Oct 15

Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening

The paper introduces Scalable Power Sampling, a training‑free method that sharpens a large language model’s generative distribution by approximating the autoregressive power distribution with a token‑level scaled low‑temperature distribution. This approach eliminates the need for iterative Markov chain Monte Carlo, achieving performance comparable to or better than one‑shot GRPO on math, QA, and code tasks while cutting inference latency by more than tenfold. The method relies solely on the base model’s outputs, requiring no external rewards or verifiers.

By Xiaotong Ji, Rasul Tutunov, Matthieu Zimmer, Haitham Bou Ammar
llmsreinforcement-learningefficiency
More like this →