AlignCP is proposed, a framework that reconciles supervised few-shot adaptation with conformal uncertainty estimation under non-exchangeability and learns a reweighted calibration distribution that reduces the score-level discrepancy between the labeled support set and the unlabeled query set.
Abstract
Uncertainty estimation for medical vision--language models (VLMs) using conformal prediction has gained increasing attention due to its distribution-free coverage guarantees. However, standard conformal prediction relies on exchangeability between calibration and test data and typically requires a sufficiently large calibration set to obtain reliable coverage. These assumptions are difficult to satisfy in few-shot transfer settings, where only a small labeled support set is available to adapt a pretrained VLM to a new medical task, while an unlabeled query set is used for evaluation. Supervised fine-tuning on the support set changes the model parameters and consequently shifts the nonconformity score distribution, breaking exchangeability between calibration and query samples and leading to unreliable coverage under distribution shift. Existing transductive conformal adaptation methods often preserve validity by avoiding supervised updates. While this helps maintain conformal assumptions, it underutilizes the scarce labeled support data and limits task adaptation, which is the primary objective in few-shot learning. In this setting, conformal prediction should serve as an uncertainty estimation layer that supports the adapted model, rather than preventing adaptation itself. To this end, we propose AlignCP, a framework that reconciles supervised few-shot adaptation with conformal uncertainty estimation under non-exchangeability. AlignCP learns a reweighted calibration distribution that reduces the score-level discrepancy between the labeled support set and the unlabeled query set. By aligning the one-dimensional nonconformity score distributions, AlignCP aims to close the coverage gap induced by adaptation without requiring query labels.
This work proposes an end-to-end framework that integrates CRM into the SSDA training objective, enabling effective CRM in the limited-labeled-target-data regime, and utilizes Optimal Transport to generate pseudolabels for unlabeled target instances.
Manos Giannopoulos, Yi Shen, Michael M. Zavlanos· 0 citations
This work systematically examines how MedSAM generalizes across diverse medical imaging benchmarks, with six adaptation strategies: full-model and encoder-only LoRA, shallow and deep visual prompt tuning (VPT), and decoder-only and full fine-tuning, and concludes that robust MedSAM adaptation requires the combined cons...
Marko Haralović, Sounic Akkaraju, Carlo Baretta et al.· 1 citation· ⚡1
The underlying causes of miscalibration in VLMs and Medical-VLMs are investigated and a simple train-time calibration method, Multi-Class Margin (MCM) regularization, is proposed, which achieves lowest ECE on 10 out of 12 settings in in-domain and remains competitive under domain shifts.
Targeted Full Conformal Prediction (T-FCP) is introduced, which uses a lightweight inductive conformal predictor to prune unlikely labels and applies FCP only to the remaining candidates, reducing computation while retaining the formal guarantee of the combined conformal procedure.
Julio Silva-Rodríguez, E. Konukoglu· arXiv.org· 0 citations
This review consolidates the landscape of CP adaptations for MLL under a unified framework, examining the types of outputs and guarantees they provide, where label dependencies are incorporated, and how inference cost scales with the number of labels.
DPQ is introduced, a lightweight pre-quantization recipe family that uses full-precision predictions to construct target-aligned calibration mixtures of high-doubt examples and generic anchors that better preserve broad multiple-choice QA behavior.
Zhen Yang, Sizai Hou, Kai-Wen Zheng et al.· 1 citation
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduNov 20, 2026
Senior Research Scientist Jim Glass discusses how his MIT-IBM Computing Research Lab collaboration connects perception and language, and addresses the shortcoming of current models.
MIT News · Artificial Intelligence· news.mit.eduNov 20, 2026
Enforcing physics-based constraints prevents AI models from generating unreliable predictions for complex nonlinear systems in weather, climate, and turbulent flows.
An MIT and Dana-Farber Institute-led team used large language models to examine nearly 4,000 studies and hundreds of code repositories, revealing gaps in the information needed to scrutinize and reuse published research.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.