# Open Problems in Mechanistic Interpretability

arXiv: https://arxiv.org/abs/2501.16496 · HTML: https://mlsanity.com/paper/2501.16496/

## Key facts

- Cited by 179 papers in the map (data through 14 Oct 2026).
- Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability.
- First posted on arXiv 27 Jan 2025.

## Placed in

- Mechanistic interpretability / [Circuits and mechanistic analysis](https://mlsanity.com/mechinterp/circuits-and-mechanistic-analysis/)

## Built on by (most cited first, 30 of 179)

- [Agentic Large Language Models, a survey](https://mlsanity.com/paper/2503.23037/) (2025)
- [MIB: A Mechanistic Interpretability Benchmark](https://mlsanity.com/paper/2504.13151/) (2025)
- [The Non-Linear Representation Dilemma: Is Causal Abstraction Enough for Mechanistic Interpretability?](https://mlsanity.com/paper/2507.08802/) (2025)
- [Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning](https://mlsanity.com/paper/2504.02922/) (2025)
- [Do Sparse Autoencoders Capture Concept Manifolds?](https://mlsanity.com/paper/2604.28119/) (2026)
- [Training Language Models to Explain Their Own Computations](https://mlsanity.com/paper/2511.08579/) (2025)
- [There Will Be a Scientific Theory of Deep Learning](https://mlsanity.com/paper/2604.21691/) (2026)
- [How Visual Representations Map to Language Feature Space in Multimodal LLMs](https://mlsanity.com/paper/2506.11976/) (2025)
- [Language Model Circuits Are Sparse in the Neuron Basis](https://mlsanity.com/paper/2601.22594/) (2026)
- [Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms](https://mlsanity.com/paper/2505.20322/) (2025)
- [How Do LLMs Acquire New Knowledge? A Knowledge Circuits Perspective on Continual Pre-Training](https://mlsanity.com/paper/2502.11196/) (2025)
- [Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis](https://mlsanity.com/paper/2505.11581/) (2025)
- [Transferring Linear Features Across Language Models With Model Stitching](https://mlsanity.com/paper/2506.06609/) (2025)
- [The Rogue Scalpel: Activation Steering Compromises LLM Safety](https://mlsanity.com/paper/2509.22067/) (2025)
- [Qwen-Scope: Turning Sparse Features into Development Tools for Large Language Models](https://mlsanity.com/paper/2605.11887/) (2026)
- [Position: Use Sparse Autoencoders to Discover Unknowns](https://mlsanity.com/paper/2506.23845/) (2025)
- [Risk Analysis Techniques for Governed LLM-based Multi-Agent Systems](https://mlsanity.com/paper/2508.05687/) (2025)
- [ProxySPEX: Inference-Efficient Interpretability via Sparse Feature Interactions in LLMs](https://mlsanity.com/paper/2505.17495/) (2025)
- [Because we have LLMs, we Can and Should Pursue Agentic Interpretability](https://mlsanity.com/paper/2506.12152/) (2025)
- [Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information](https://mlsanity.com/paper/2502.14258/) (2025)
- [Alternating Gradient Flows: A Theory of Feature Learning in Two-layer Neural Networks](https://mlsanity.com/paper/2506.06489/) (2025)
- [I Predict Therefore I Am: Is Next Token Prediction Enough to Learn Human-Interpretable Concepts from Data?](https://mlsanity.com/paper/2503.08980/) (2025)
- [RelP: Faithful and Efficient Circuit Discovery in Language Models via Relevance Patching](https://mlsanity.com/paper/2508.21258/) (2025)
- [Distributional AGI Safety](https://mlsanity.com/paper/2512.16856/) (2025)
- [Do Sparse Autoencoders Identify Reasoning Features in Language Models?](https://mlsanity.com/paper/2601.05679/) (2026)
- [Coding Agents are Effective Long-Context Processors](https://mlsanity.com/paper/2603.20432/) (2026)
- [The Information Geometry of Softmax: Probing and Steering](https://mlsanity.com/paper/2602.15293/) (2026)
- [Formal Mechanistic Interpretability: Automated Circuit Discovery with Provable Guarantees](https://mlsanity.com/paper/2602.16823/) (2026)
- [Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Framework](https://mlsanity.com/paper/2506.15538/) (2025)
- [The Singapore Consensus on Global AI Safety Research Priorities](https://mlsanity.com/paper/2506.20702/) (2025)

## Abstract (arXiv)

Mechanistic interpretability aims to understand the computational mechanisms underlying neural networks' capabilities in order to accomplish concrete scientific and engineering goals. Progress in this field thus promises to provide greater assurance over AI system behavior and shed light on exciting scientific questions about the nature of intelligence. Despite recent progress toward these goals, there are many open problems in the field that require solutions before many scientific and practical benefits can be realized: Our methods require both conceptual and practical improvements to reveal deeper insights; we must figure out how best to apply our methods in pursuit of specific goals; and the field must grapple with socio-technical challenges that influence and are influenced by our work. This forward-facing review discusses the current frontier of mechanistic interpretability and the open problems that the field may benefit from prioritizing.