When Trees Are Not Enough: Learning Mixed-Topology Feature Graphs with Adaptive Graph Sparse Autoencoders
cs.LG, cs.CL
Submitted: 2026-09-28
Updated: 2026-09-28
Terminology
Sources
- BatchTopK Sparse Autoencoders
- Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders
- Feature Hedging: Correlated Features Break Narrow Sparse Autoencoders
- Subspace-Aware Sparse Autoencoders for Effective Mechanistic Interpretability
- Gemma 2: Improving Open Language Models at a Practical Size
- Do Sparse Autoencoders Learn Meaningful Concept Hierarchies?
- Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders
- C$^{2}$R: Cross-sample Consistency Regularization Mitigates Feature Splitting and Absorption in Sparse Autoencoders
- The MiniPile Challenge for Data-Efficient Language Models
- OrtSAE: Orthogonal Sparse Autoencoders Uncover Atomic Features
- From Atoms to Trees: Building a Structured Feature Forest with Hierarchical Sparse Autoencoders
- k-Sparse Autoencoders
- Incorporating Hierarchical Semantics in Sparse Autoencoder Architectures
- The Geometry of Categorical and Hierarchical Concepts in Large Language Models
- Qwen3 Technical Report
- Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks