Concept-based abductive and contrastive explanations find minimal high-level concepts that causally determine vision model outcomes on individual images or groups sharing a specified behavior.
Concept bottleneck models
9 Pith papers cite this work, alongside 151 external citations. Polarity classification is still indexing.
9
Pith papers citing it
151
external citations · Pith
abstract
We seek to learn models that we can interact with using high-level concepts: if the model did not think there was a bone spur in the x-ray, would it still predict severe arthritis? State-of-the-art models today do not typically support the manipulation of concepts like "the existence of bone spurs", as they are trained end-to-end to go directly from raw input (e.g., pixels) to output (e.g., arthritis severity). We revisit the classic idea of first predicting concepts that are provided at training time, and then using these concepts to predict the label. By construction, we can intervene on these concept bottleneck models by editing their predicted concept values and propagating these changes to the final prediction. On x-ray grading and bird identification, concept bottleneck models achieve competitive accuracy with standard end-to-end models, while enabling interpretation in terms of high-level clinical concepts ("bone spurs") or bird attributes ("wing color"). These models also allow for richer human-model interaction: accuracy improves significantly if we can correct model mistakes on concepts at test time.
citation-role summary
background 1
citation-polarity summary
years
2026 9roles
background 1polarities
background 1representative citing papers
Introduces synthetic benchmarks for concept bottleneck models that control data modality, concept choice, annotation quality, and completeness to evaluate performance in decision support and automation.
TimeSRL uses semantic abstractions from time-series data optimized via reinforcement learning to achieve better cross-dataset generalization than standard ML or LLM baselines in mental health prediction.
An attribute-guided dual-branch framework fuses a standard classifier with an interpretable attribute-prior branch to boost ultrasound classification accuracy and explainability.
VP2O maps PPO to SVGD in a MoE architecture using functional kernels and expert orthogonalization, claiming +179 ELO on Codeforces and 32% token reduction on AIME for a 33B/4B model.
CLEAR-HPV restructures the latent space of attention-based MIL models to discover 10 label-free morphologic concepts that preserve slide-level HPV prediction performance and generalize across TCGA-HNSCC, TCGA-CESC, and CPTAC-HNSCC datasets.
QuAP is a working prototype combining similarity-based audio retrieval, real-time procedural models, and a perceptually guided parameter assistant, with evaluations showing quality gains in five of six synthesis models and positive feedback from 16 practitioners.
A position paper argues that post-hoc XAI explanations are unfaithful and paradoxical, proposing a shift to expert-based verification and certification of AI systems.
The paper introduces 'Platonic Projection Structures,' a reformulation of standard PSD operator theory applied to representation learning, with experiments that verify definitions rather than test predictions.
citing papers explorer
-
Concept-Based Abductive and Contrastive Explanations for Behaviors of Vision Models
Concept-based abductive and contrastive explanations find minimal high-level concepts that causally determine vision model outcomes on individual images or groups sharing a specified behavior.
-
Measuring What Matters: Synthetic Benchmarks for Concept Bottleneck Models
Introduces synthetic benchmarks for concept bottleneck models that control data modality, concept choice, annotation quality, and completeness to evaluate performance in decision support and automation.
-
TimeSRL: Generalizable Time-Series Behavioral Modeling via Semantic RL-Tuned LLMs -- A Case Study in Mental Health
TimeSRL uses semantic abstractions from time-series data optimized via reinforcement learning to achieve better cross-dataset generalization than standard ML or LLM baselines in mental health prediction.
-
Boosting Ultrasound Image Classification via Attribute-Guided Dual-Branch Framework
An attribute-guided dual-branch framework fuses a standard classifier with an interpretable attribute-prior branch to boost ultrasound classification accuracy and explainability.
-
Variational Proximal Policy Optimization
VP2O maps PPO to SVGD in a MoE architecture using functional kernels and expert orthogonalization, claiming +179 ELO on Codeforces and 32% token reduction on AIME for a 33B/4B model.
-
CLEAR-HPV: Interpretable concept discovery for human-papillomavirus-associated morphology in whole-slide histology
CLEAR-HPV restructures the latent space of attention-based MIL models to discover 10 label-free morphologic concepts that preserve slide-level HPV prediction performance and generalize across TCGA-HNSCC, TCGA-CESC, and CPTAC-HNSCC datasets.
-
Quality Audio Prototyping: a prototype system for unified sound retrieval and procedural generation
QuAP is a working prototype combining similarity-based audio retrieval, real-time procedural models, and a perceptually guided parameter assistant, with evaluations showing quality gains in five of six synthesis models and positive feedback from 16 practitioners.
-
Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions
A position paper argues that post-hoc XAI explanations are unfaithful and paradoxical, proposing a shift to expert-based verification and certification of AI systems.
-
Platonic Projection Structures: Operator-Induced Observability in Representation Learning
The paper introduces 'Platonic Projection Structures,' a reformulation of standard PSD operator theory applied to representation learning, with experiments that verify definitions rather than test predictions.