Federated deepfake Grad-CAM faithfulness collapses at extreme heterogeneity
An EfficientNet-B4 on FaceForensics++ matches a centralized baseline until client data is extremely non-IID, when explanations turn misleading.
Stories per day · 30 days
Relatedmultimodal2attribution1catastrophic-forgetting1computer-vision1concept-learning1concrete1continual-learning1counterfactual-explanations1counterfactual-training1crane1credit1deepfake1document-understanding1faceforensics1federated-learning1grad-cam1gradient-boosting1language-models1
24 stories, newest first · one card per story
Stories tagged #explainability
DocXEval benchmarks explainability methods for multimodal document models
Occlusion ranked most faithful and interpretable across 13 methods, while some metrics were unreliable.
CREDIT training improves language models’ self-attribution of sources
Counterfactual labels on the final answer beat prompting and reasoning-level fine-tuning on four open models, including two distilled reasoners.
Concept-consistent continual learning stops saliency explanation drift
C3L anchors regularization to offline class concepts, and authors report gains on five benchmarks against over 20 baselines.
CRANE retrieves real examples as multimodal counterfactual explanations
Multimodal Consensus Retrieval was best or joint-best on three of four datasets across three fusion strategies.
Ensemble models predict heat-exposed concrete strength at R² 0.903
On 1,204 records, gradient boosting had the lowest error at 6.67 MPa RMSE and was packaged in a Python GUI.
Spectral-Spatial Interpretation attributes images by position and channel
On HLS Burn Scar, SSI beat an MC-Owen estimator while using about 6% of its evaluation budget.
ConEx: Human-Interpretable Saliency Maps via Concept-Aware Attribution
ConEx ties vision saliency maps to automatically discovered concepts using unsupervised concept activation vectors.
MEA: A Reward-Driven Multi-Agent System for Faithful Model Explanations
MEA trains a multi-agent system to produce more faithful natural-language explanations across tabular, text, and vision models.
Explaining Hyperbolic Neural Networks via Geometry-Aware Relevance Propagation
LRP-radial-all explains hyperbolic networks with attributions consistent across equivalent geometric implementations.
DF-CBM: Region-Aware Concept Bottleneck Models for Deepfake Detection
DF-CBM is a region-aware concept bottleneck model that explains deepfake decisions with spatially grounded manipulation concepts.
Uncertainty and Explainability in Deep Rough Volatility: A Neural Information-Theoretic Posterior Approach
Neural posterior inference calibrates rough Heston models and uses Hellinger-SHAP to explain implied-volatility information gain.
X-SPUR: Explainable Surprisal-Based Protocol-Aware Unsupervised Reasoning for Automotive Ethernet Intrusion Detection
X-SPUR flags Automotive Ethernet attacks unsupervised via token surprisal, scoring AUC 0.9987 on TOW-IDS.
A Qualitative Model for Reasoning about Path and Support
A hybrid qualitative reasoning model with center-of-mass stability logic enables explainable path-planning and stability guidance for the Camelot Jr. block-puzzle game.
Probabilistic Linear Explanations
Researchers introduce a unified probabilistic explainability framework using sparse anchored linear models that outperforms LIME and MAPLE on relevance error.
Det-LIME: Detector-Aware, Multi-Instance Local Interpretable Model-Agnostic Explanations for Automated Marine Mammal Detection
Det-LIME extends LIME to multi-instance object detection explanations, improving attribution for harbor seal aerial surveys.
Recurrent GraphNeural NetworkswithSet-BasedAggregation
Paper proves two-directional equivalence between recurrent GNNs with set-based aggregation and Boolean closure of reachability/safety properties in modal mu-calculus, checkable from weights.
DynSHAP: Towards Explainable Dynamic Survival Analysis
DynSHAP extends SHAP explainability to dynamic survival analysis, treating time-feature pairs as Shapley players for longitudinal clinical predictions.
Explainability Assistant: A Conversational XAI Interface for Interpreting Energy Consumption Models
Researchers release Explainability Assistant, an open-source conversational XAI tool using LLM function calling, lifting intent-parsing accuracy from 76.8% to 94%.
Understanding the Impact of Model Pruning on Long-Tail Forgetting and Explanation Reliability in Medical Imaging
Systematic study finds model pruning causes frequency-dependent long-tail forgetting in medical imaging and that gradient-informed methods best preserve explanations.
Bag of Tricks or Bag of Myths? Reducing Modeling Complexity with Task Knowledge in Explainable Suicide Risk Assessment
Researchers audit 31 NLP techniques for clinician-annotated suicide risk prediction, finding only 5 of 31 comparisons yield reliable gains.
CARDEA: Auditable Reasoning Grounded in Spatial Evidence for End-to-End Coronary Angiography Interpretation
CARDEA, a vision-language model trained only on public data, matches cardiologists on coronary angiography complexity assessment while exposing auditable bounding-box evidence.
Why Is SHAP Not a Reliable Standalone Explanation Framework for Malware Detection?
arXiv paper shows SHAP gives unreliable standalone explanations for malware detection, with attribution dilution and sign reversal in dependent PE feature spaces.
AI for Military Support
Study of 2,015 Israeli military personnel found algorithmic aversion toward AI targeting decision support, reduced when explainable AI features were added.