When Fine-Tuning Changes the Evidence: Architecture-Dependent Semantic Drift in Chest X-Ray Explanations

• URL: http://arxiv.org/abs/2604.08513v1
• Date: Thu, 09 Apr 2026 17:53:02 GMT
• Title: When Fine-Tuning Changes the Evidence: Architecture-Dependent Semantic Drift in Chest X-Ray Explanations
• Abstract summary: Transfer learning followed by fine-tuning is widely adopted in medical image classification due to consistent gains in diagnostic performance.<n>In multi-class settings with overlapping visual features, improvements in accuracy do not guarantee stability of the visual evidence used to support predictions.<n>We define semantic drift as systematic changes in the attribution structure supporting a model's predictions between transfer learning and full fine-tuning.
• Score: 5.127836966483091
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Transfer learning followed by fine-tuning is widely adopted in medical image classification due to consistent gains in diagnostic performance. However, in multi-class settings with overlapping visual features, improvements in accuracy do not guarantee stability of the visual evidence used to support predictions. We define semantic drift as systematic changes in the attribution structure supporting a model's predictions between transfer learning and full fine-tuning, reflecting potential shifts in underlying visual reasoning despite stable classification performance. Using a five-class chest X-ray task, we evaluate DenseNet201, ResNet50V2, and InceptionV3 under a two-stage training protocol and quantify drift with reference-free metrics capturing spatial localization and structural consistency of attribution maps. Across architectures, coarse anatomical localization remains stable, while overlap IoU reveals pronounced architecture-dependent reorganization of evidential structure. Beyond single-method analysis, stability rankings can reverse across LayerCAM and GradCAM++ under converged predictive performance, establishing explanation stability as an interaction between architecture, optimization phase, and attribution objective.

Related papers

• Orthogonal JEPA: Factorized Predictive States for Latent World Models [95.20158869896393]
  We introduce method, a latent world-modeling framework based on predictive factorization.<n>We show that method can allocate redundant capacity to dominant signals while providing weak or conflicting gradients to less dominant predictive structure.<n>Experiments on controlled vision, single-cell transcriptomics, longitudinal health records, continuous control, and molecular dynamics evaluate representation quality, forecasting, planning, and long-horizon stability.
  arXiv  Detail & Related papers  (2026-08-20T13:59:57Z)
• When Geometry Aligns: Dihedral Hidden-State Transformations in UNet, ViT, and DiT Architectures [20.803275645784694]
  We study the effects of structured geometric perturbations within Diffusion Transformers (DiTs)<n>We find that consistent transformations improve stability, while inconsistent ones induce predictable, architecture-specific failures.<n> Taken together with supporting ViT and controlled DiT analyses, these results establish geometric consistency as a key principle for stable hidden-state interventions.
  arXiv  Detail & Related papers  (2026-07-03T19:47:07Z)
• Hierarchical Consistency Learning for Test-time Adaptation in Camouflage Perception [50.278200968044665]
  Camouflaged object detection (COD) aims to localize targets that exhibit minimal perceptual differences from backgrounds through physical attributes.<n>Existing methods, constrained by the static train-then-freeze paradigm, suffer from domain rigidity and dependency annotation.<n>We propose the hierarchical consistency learning framework, which integrates test-time adaptation for dynamic representation recalibration.
  arXiv  Detail & Related papers  (2026-05-25T09:57:46Z)
• Beyond Point-Wise Matching: Structural Representation Alignment for Accelerating Diffusion Transformers [93.3976834364707]
  We propose sREPA, a structural REPresentation Alignment framework to enforce consistency in the relational geometry of feature maps.<n>By encouraging the model to internalize holistic spatial layouts and structural correlations from pre-trained features, sREPA achieves faster and more stable convergence.
  arXiv  Detail & Related papers  (2026-05-16T12:01:04Z)
• A Persistent Homology Design Space for 3D Point Cloud Deep Learning [5.810033007426438]
  Persistent Homology (PH) offers stable, multi-scale descriptors of intrinsic shape structure.<n>Despite strong theoretical guarantees and increasing empirical adoption, its integration into deep learning for point clouds remains largely ad hoc and architecturally peripheral.<n>In this work, we introduce a unified design space for Persistent-Homology driven learning in 3D point clouds.
  arXiv  Detail & Related papers  (2026-04-05T22:38:37Z)
• StepVAR: Structure-Texture Guided Pruning for Visual Autoregressive Models [98.72926158261937]
  We propose a training-free token pruning framework for Visual AutoRegressive models.<n>We employ a lightweight high-pass filter to capture local texture details, while leveraging Principal Component Analysis (PCA) to preserve global structural information.<n>To maintain valid next-scale prediction under sparse tokens, we introduce a nearest neighbor feature propagation strategy.
  arXiv  Detail & Related papers  (2026-03-02T11:35:05Z)
• Invariance on Manifolds: Understanding Robust Visual Representations for Place Recognition [19.200074425090595]
  We propose a Second-Order Geometric Statistics framework that inherently captures geometric stability without training.<n>Our approach introduces a training-free framework built upon fixed, pre-trained backbones, achieving strong zero-shot generalization without parameter updates.
  arXiv  Detail & Related papers  (2026-01-31T18:12:29Z)
• Cross-Domain Transfer with Self-Supervised Spectral-Spatial Modeling for Hyperspectral Image Classification [5.784164305429653]
  This paper proposes a self-supervised cross-domain transfer framework.<n>It learns transferable spectral-spatial joint representations without source labels.<n> Experimental results demonstrate stable classification performance and strong cross-domain adaptability.
  arXiv  Detail & Related papers  (2026-01-26T02:52:35Z)
• ALOcc: Adaptive Lifting-Based 3D Semantic Occupancy and Cost Volume-Based Flow Predictions [91.55655961014027]
  3D semantic occupancy and flow prediction are fundamental to understanding scene scene.<n>This paper proposes a vision-based framework with three targeted improvements.<n>Our purely convolutional architecture establishes new SOTA performance on multiple benchmarks for both semantic occupancy and joint semantic-flow prediction.
  arXiv  Detail & Related papers  (2024-11-12T11:32:56Z)
• Intensity Profile Projection: A Framework for Continuous-Time Representation Learning for Dynamic Networks [50.2033914945157]
  We present a representation learning framework, Intensity Profile Projection, for continuous-time dynamic network data. The framework consists of three stages: estimating pairwise intensity functions, learning a projection which minimises a notion of intensity reconstruction error. Moreoever, we develop estimation theory providing tight control on the error of any estimated trajectory, indicating that the representations could even be used in quite noise-sensitive follow-on analyses.
  arXiv  Detail & Related papers  (2023-06-09T15:38:25Z)
• Motion-Scenario Decoupling for Rat-Aware Video Position Prediction: Strategy and Benchmark [49.58762201363483]
  We introduce RatPose, a bio-robot motion prediction dataset constructed by considering the influence factors of individuals and environments. We propose a Dual-stream Motion-Scenario Decoupling framework that effectively separates scenario-oriented and motion-oriented features. We demonstrate significant performance improvements of the proposed textitDMSD framework on different difficulty-level tasks.
  arXiv  Detail & Related papers  (2023-05-17T14:14:31Z)
• Robust Self-Supervised LiDAR Odometry via Representative Structure Discovery and 3D Inherent Error Modeling [67.75095378830694]
  We develop a two-stage odometry estimation network, where we obtain the ego-motion by estimating a set of sub-region transformations. In this paper, we aim to alleviate the influence of unreliable structures in training, inference and mapping phases. Our two-frame odometry outperforms the previous state of the arts by 16%/12% in terms of translational/rotational errors.
  arXiv  Detail & Related papers  (2022-02-27T12:52:27Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.