aiqbench analysis
Evidence Field introduces a novel spatial-program architecture for medical imaging that enables the independent reconstruction of decision computations alongside class scores. By jointly generating covering programs and signed evidence, this approach addresses the lack of native interpretability in standard classifiers while maintaining rigorous mathematical bounds on inference. The work represents a significant step toward transparent AI systems in high-stakes domains like histopathology, where understanding the spatial logic behind predictions is critical.
Evidence Field: Adaptive Covering Programs for Medical Imaging
arXiv:2610.24773v1 Announce Type: cross Abstract: Medical image classifiers typically return predictions without a native contract for independently reconstructing their spatial decision computation. Evidence Field is a trainable spatial-program architecture that jointly produces class scores, selected covering programs and signed evidence. Its encoder combines absolute and image-relative response fields with associative spatial moments and learned characteristic sums. A typed classifier scores individual regions and parent-conditioned multiplicative groups; a covering recursion selects scale, partition and composition. We establish exact one-fold coverage, optimality within the registered grammar, a bound relating smooth training to hard inference, signed-score conservation and class-centred offset invariance. Direct contractions and their adjoint specify arithmetic and storage scaling on two- and three-dimensional grids. Native output records bind inputs, parameters, supports, types and contextual dependencies to individual and ensemble predictions. In a separately reported histopathology application, patient-separated training and evaluation across held-out centres provide empirical evidence for the architecture. Trained mechanism and corrected-loss comparisons distinguish predictive effects from implementation correctness; independent readers reconstruct descriptors, routes, scores and probability mixtures. The principal contribution is a coupled representation, decision and evidence architecture whose selected computation can be reproduced independently.
Abstract:Medical image classifiers typically return predictions without a native contract for independently reconstructing their spatial decision computation. Evidence Field is a trainable spatial-program architecture that jointly produces class scores, selected covering programs and signed evidence. Its encoder combines absolute and image-relative response fields with associative spatial moments and learned characteristic sums. A typed classifier scores individual regions and parent-conditioned multiplicative groups; a covering recursion selects scale, partition and composition. We establish exact one-fold coverage, optimality within the registered grammar, a bound relating smooth training to hard inference, signed-score conservation and class-centred offset invariance. Direct contractions and their adjoint specify arithmetic and storage scaling on two- and three-dimensional grids. Native output records bind inputs, parameters, supports, types and contextual dependencies to individual and ensemble predictions. In a separately reported histopathology application, patient-separated training and evaluation across held-out centres provide empirical evidence for the architecture. Trained mechanism and corrected-loss comparisons distinguish predictive effects from implementation correctness; independent readers reconstruct descriptors, routes, scores and probability mixtures. The principal contribution is a coupled representation, decision and evidence architecture whose selected computation can be reproduced independently.
Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2610.24773 [eess.IV]
(or arXiv:2610.24773v1 [eess.IV] for this version)
https://doi.org/10.48550/arXiv.2610.24773 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 289 words
Related coverage
Faithful Chart Generation for Multimodal Deep Research: Frame-Evidence Co-Adaptation
arXiv · Oct 2, 2026
Backbone-Adaptive Evidence Routing for Robust Pairwise LLM Judging
arXiv · Sep 28, 2026
ReLATE: Reliability-Guided Evidence Fusion for Robust UAV--Satellite cross-view Geo-Localization
arXiv · Jul 29, 2026
A Tree-based RAG Framework for Evidence-Intensive QA via Adaptive Planning and Topology-Aware Evidence Gathering
arXiv · Sep 7, 2026