cs.LGNov 16, 2026

One-Shot Calibration via Model Editing

Authors: Jiaxing Qiu, Luke Hakso, Roxana Daneshjou, Ahmed Alaa, Thomas Hartvigsen

Organizations: University of Virginia · Stanford University · University of California, Berkeley

Abstract

Model editing aims to correct errors in large, pretrained models by replacing model generations with user-provided responses. Existing editors assume a single target answer for each edit. However, many real tasks inherently entail uncertainty, resulting in multiple plausible answers with varying likelihoods in user feedback. Such feedback forms a target distribution to not only correct but also calibrate model generations. We therefore propose SoftEdit, the first model editor to leverage uncertainty in user responses and perform one-shot calibration during editing. To generalize calibration to unseen samples, SoftEdit uses co-plausibility relationships between answers while preserving real distributional shifts, balancing the new distributional generality-locality tradeoff. Across three language models, three vision-language models, and two large datasets, SoftEdit achieves state-of-the-art performance, ultimately showing that uncertainty in user responses enables generalizable calibration of pretrained models beyond single-answer corrections.

Figures & tables

Explore similar work

CardsList
  1. HeadEdit: Calibrating Language Model Behavior Through the Frozen Unembedding Matrix

    Oct 1, 2026Zirui He, Haiyan Zhao, Jingyu Hu +5Language Model SteeringLanguage Model Calibration

  2. Generalizable Model Editing Could Be More Than Direct Generalization

    Oct 11, 2026Xiaotian Ye, Chunyao Yang, Xiaohan Wang +4Model EditingLanguage Model Editing

  3. CrispEdit: Low-Curvature Projections for Scalable Non-Destructive LLM Editing

    Feb 17, 2026Zarif Ikram, Arad Firouzkouhi, Stephen Tu +2Language Model Editing