From Input to Output: A Flexible Agent for Dual-End Interpretation of Sparse Autoencoder Features

arXiv:2609.35367 · cs.CL · Submitted 2026-09-28 · Read on arXiv

cs.CL

Submitted: 2026-09-28

Updated: 2026-09-28

Code: https://github.com/THUAIS-Lab/DAFI

Terminology

• Steering — 13× in this paper · explained in 2 episodes
• Trace — 10× in this paper · explained in 2 episodes
• Consistency — 7× in this paper · explained in 3 episodes
• Interpretability — 6× in this paper · explained in 15 episodes
• LLM judge — 4× in this paper · explained in 2 episodes
• Core — 3× in this paper · explained in 2 episodes
• Robustness — 3× in this paper · explained in 5 episodes
• Sparse Autoencoders (SAEs) — 3× in this paper · explained in 3 episodes
• Feature Selection — 2× in this paper · explained in 4 episodes
• Iterative Refinement — 2× in this paper · explained in 3 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Abstraction — 1× in this paper · explained in 1 episode
• Adaptive Probing — 1× in this paper · explained in 1 episode
• Agentic System — 1× in this paper · explained in 2 episodes
• Aggregation — 1× in this paper · explained in 1 episode
• Benchmark — 1× in this paper · explained in 2 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Dual-End Agentic Feature Interpretation — 1× in this paper
• Elias — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Fidelity — 1× in this paper · explained in 2 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Mechanistic interpretability — 1× in this paper · explained in 2 episodes
• Perplexity — 1× in this paper · explained in 3 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Self-Evolving Agents — 1× in this paper · explained in 2 episodes

Sources

• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• Qwen-Scope: Turning Sparse Features into Development Tools for Large Language Models
• Toy Models of Superposition
• Gemma 2: Improving Open Language Models at a Practical Size
• Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2
• Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
• Qwen3 Technical Report
• SkillWeaver: Web Agents can Self-Improve by Discovering and Honing Skills

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering