ICLR 2026 Past Other
ICLR 2026 Workshop on Unifying Concept Representation Learning
UCRL@ICLR2026
- Submission deadline
- Feb 4, 2026, 12:00 UTC imported from OpenReview — check the website for extensions
- Submission portal
- OpenReview
- Notes
- Auto-imported from the OpenReview venue record on 2026-06-10 — please verify and enrich (topics are keyword-guessed).
Accepted papers (26)
Fetched from OpenReview (v2) on 2026-06-10.
-
An Information-Theoretic Approach to Benign Leakage in Static Concept Embedding Models
-
Automated Concept Discovery for LLM-as-a-Judge Preference Analysis
-
Automatic Interpretation of Visual Concepts
-
Causality Guided Representation Learning for Cross-Style Hate Speech Detection
-
Concept Bottleneck Diffusion for Steerable Generation
-
Continual learning and refinement of causal models through dynamic predicate invention
-
Decomposing Representation Drift via Interventions
-
Diagnosing and Fixing Latent Recovery in Sparse Autoencoders
-
Digging Deeper: Learning Multi-Level Concept Hierarchies
-
Do Monolingual Language Models Learn Cross-Lingual Universal Conceptual Representations?
-
Dynamics Reveals Structure: Challenging the Linear Propagation Assumption
-
Extracting Representations in LLMs Robust to Distribution Shifts
-
Gluing Local Contexts into Global Meaning: A Sheaf-Theoretic Decomposition of Transformer Representations
-
Identifiable Estimation of Causal Concept Effects under Visual Latent Confounding
-
Multi-shot AutoInterp: Agents Can Explain Complex Features By Refining Explanations
-
Navigating the Concept Space of Language Models
-
NEUROLOGIC: From Neural Representations to Interpretable Logic Rules
-
On the Existence and Behavior of Secondary Attention Sinks
-
One Language, Two Scripts: Probing Script-Invariance in LLM Concept Representations
-
Post-hoc Self-explanation of CNNs
-
Post-hoc Stochastic Concept Bottleneck Models
-
Self-Interpretable Concept Representations: Training Lightweight Adapters on Vector-Label Pairs
-
Simulating Concept Bottlenecks with Vision-Language Models
-
SSCD: Sparse Semantic Concept Defense Against Semantic Adversarial Attacks to Image Classifiers
-
Syntax-Preserving Hyperbolic Visual-Semantic Embeddings
-
VLC Fusion: Vision-Language Conditioned Sensor Fusion for Robust Object Detection