DUNE: Disentangling Polysemantic Neurons Without Losing Predictive Performance
cs.LG, cs.CV
Submitted: 2026-05-29
Updated: 2026-11-11
Terminology
Sources
- On the Opportunities and Risks of Foundation Models
- Diffusion-Pretrained Dense and Contextual Embeddings
- A Survey on Hallucination in Large Vision-Language Models
- The Quest for the Right Mediator: Surveying Mechanistic Interpretability Through the Lens of Causal Mediation Analysis
- A Practical Review of Mechanistic Interpretability for Transformer-Based Language Models
- Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
- Polysemanticity and Capacity in Neural Networks
- Gemma 3 Technical Report
- Qwen3 Technical Report
- SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features
- Steering Language Models With Activation Engineering
- CFM: Language-aligned Concept Foundation Model for Vision
- InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis