Cross-Layer TopK: Dynamically Allocating Sparsity Across Layers in Transcoders
cs.LG
Submitted: 2026-11-06
Updated: 2026-11-06
Terminology
Sources
- Adaptive Sparse Allocation with Mutual Choice & Feature Choice Sparse Autoencoders
- CLT-Forge: A Scalable Library for Cross-Layer Transcoders and Attribution Graphs
- Decoupled Weight Decay Regularization
- Olmo 3
- Transcoders Beat Sparse Autoencoders for Interpretability
- Improving Dictionary Learning with Gated Sparse Autoencoders
- Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
- Qwen3 Technical Report
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