disentangled representations
Disentangled representations are learned features that separate different underlying factors or variables in the data, facilitating better generalization and interpretability in machine learning models.
- CG-SSL: Concept-Guided Self-Supervised Learning
- Can Diffusion Models Disentangle? A Theoretical Perspective
- Causal Differentiating Concepts: Interpreting LM Behavior via Causal Representation Learning
- Disentanglement Beyond Static vs. Dynamic: A Benchmark and Evaluation Framework for Multi-Factor Sequential Representations
- Learning Interactive World Model for Object-Centric Reinforcement Learning
- Mind-the-Glitch: Visual Correspondence for Detecting Inconsistencies in Subject-Driven Generation
- Preventing Shortcuts in Adapter Training via Providing the Shortcuts
- scMRDR: A scalable and flexible framework for unpaired single-cell multi-omics data integration