{"total":77,"rows":[{"id":"2404.14082","title":"Mechanistic Interpretability for AI Safety -- A Review","created":"2024-04-22","cited_by_count":422,"area":"mechinterp"},{"id":"2309.16042","title":"Towards Best Practices of Activation Patching in Language Models: Metrics and Methods","created":"2023-09-27","cited_by_count":359,"area":"mechinterp"},{"id":"2305.01610","title":"Finding Neurons in a Haystack: Case Studies with Sparse Probing","created":"2023-05-02","cited_by_count":334,"area":null},{"id":"2309.00941","title":"Emergent Linear Representations in World Models of Self-Supervised Sequence Models","created":"2023-09-02","cited_by_count":319,"area":"wm"},{"id":"2301.04709","title":"Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretability","created":"2023-01-11","cited_by_count":205,"area":"mechinterp"},{"id":"2605.29358","title":"Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet","created":"2026-05-28","cited_by_count":191,"area":"mechinterp"},{"id":"2407.02646","title":"A Practical Review of Mechanistic Interpretability for Transformer-Based Language Models","created":"2024-07-02","cited_by_count":118,"area":"mechinterp"},{"id":"2301.05062","title":"Tracr: Compiled Transformers as a Laboratory for Interpretability","created":"2023-01-12","cited_by_count":104,"area":null},{"id":"2403.08946","title":"Usable XAI: 10 Strategies Towards Exploiting Explainability in the LLM Era","created":"2024-03-13","cited_by_count":46,"area":null},{"id":"2408.05451","title":"Mathematical Models of Computation in Superposition","created":"2024-08-10","cited_by_count":35,"area":"mechinterp"},{"id":"2501.09929","title":"Interpretable Steering of Large Language Models with Feature Guided Activation Additions","created":"2025-01-17","cited_by_count":35,"area":"mechinterp"},{"id":"2501.05398","title":"Mechanistic understanding and validation of large AI models with SemanticLens","created":"2025-01-09","cited_by_count":29,"area":null},{"id":"2304.09707","title":"Disentangling Neuron Representations with Concept Vectors","created":"2023-04-19","cited_by_count":28,"area":"mechinterp"},{"id":"2505.22483","title":"A Closer Look at Multimodal Representation Collapse","created":"2025-05-28","cited_by_count":28,"area":null},{"id":"2407.14494","title":"InterpBench: Semi-Synthetic Transformers for Evaluating Mechanistic Interpretability Techniques","created":"2024-07-19","cited_by_count":25,"area":"mechinterp"},{"id":"2502.10835","title":"Back Attention: Understanding and Enhancing Multi-Hop Reasoning in Large Language Models","created":"2025-02-15","cited_by_count":23,"area":"lr"},{"id":"2409.15318","title":"On the Complexity of Neural Computation in Superposition","created":"2024-09-05","cited_by_count":21,"area":"mechinterp"},{"id":"2506.15963","title":"On the Limits of Sparse Autoencoders: A Theoretical Framework and Reweighted Remedy","created":"2025-06-19","cited_by_count":14,"area":"mechinterp"},{"id":"2211.09169","title":"Engineering Monosemanticity in Toy Models","created":"2022-11-16","cited_by_count":13,"area":null},{"id":"2502.14050","title":"Diversity-driven Data Selection for Language Model Tuning through Sparse Autoencoder","created":"2025-02-19","cited_by_count":12,"area":"mechinterp"},{"id":"2508.11214","title":"How Causal Abstraction Underpins Computational Explanation","created":"2025-08-15","cited_by_count":11,"area":null},{"id":"2506.15538","title":"Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Framework","created":"2025-06-18","cited_by_count":10,"area":null},{"id":"2603.09972","title":"From Data Statistics to Feature Geometry: How Correlations Shape Superposition","created":"2026-03-10","cited_by_count":10,"area":"mechinterp"},{"id":"2504.08842","title":"Towards Combinatorial Interpretability of Neural Computation","created":"2025-04-10","cited_by_count":9,"area":null},{"id":"2505.14530","title":"Internal Chain-of-Thought: Empirical Evidence for Layer-wise Subtask Scheduling in LLMs","created":"2025-05-20","cited_by_count":9,"area":"lr"},{"id":"2509.25758","title":"Thinking Sparks!: Emergent Attention Heads in Reasoning Models During Post Training","created":"2025-09-30","cited_by_count":8,"area":"lr"},{"id":"2510.11709","title":"Adversarial Attacks Leverage Interference Between Features in Superposition","created":"2025-10-13","cited_by_count":8,"area":null},{"id":"2602.20330","title":"Circuit Tracing in Vision-Language Models: Understanding the Internal Mechanisms of Multimodal Thinking","created":"2026-02-23","cited_by_count":8,"area":"lr"},{"id":"2507.09937","title":"Memorization Sinks: Isolating Memorization during LLM Training","created":"2025-07-14","cited_by_count":7,"area":null},{"id":"2512.13568","title":"Superposition as Lossy Compression: Measure with Sparse Autoencoders and Connect to Adversarial Vulnerability","created":"2025-12-15","cited_by_count":7,"area":"mechinterp"},{"id":"2310.13040","title":"Interpreting CLIP: Insights on the Robustness to ImageNet Distribution Shifts","created":"2023-10-19","cited_by_count":6,"area":null},{"id":"2501.06141","title":"Emergent Symbol-like Number Variables in Artificial Neural Networks","created":"2025-01-10","cited_by_count":6,"area":null},{"id":"2511.19265","title":"Unboxing the Black Box: Mechanistic Interpretability for Algorithmic Understanding of Neural Networks","created":"2025-11-24","cited_by_count":6,"area":"mechinterp"},{"id":"2502.14133","title":"Self-Regularization with Sparse Autoencoders for Controllable LLM-based Classification","created":"2025-02-19","cited_by_count":5,"area":"mechinterp"},{"id":"2508.15094","title":"Evaluating Sparse Autoencoders for Monosemantic Representation","created":"2025-08-20","cited_by_count":4,"area":"mechinterp"},{"id":"2511.20798","title":"Physics Steering: Causal Control of Cross-Domain Concepts in a Physics Foundation Model","created":"2025-11-25","cited_by_count":4,"area":"mechinterp"},{"id":"2601.22012","title":"Putting a Face to Forgetting: Continual Learning meets Mechanistic Interpretability","created":"2026-01-29","cited_by_count":4,"area":"mechinterp"},{"id":"2310.07837","title":"Measuring Feature Sparsity in Language Models","created":"2023-10-11","cited_by_count":3,"area":null},{"id":"2312.03051","title":"Generating Interpretable Networks using Hypernetworks","created":"2023-12-05","cited_by_count":3,"area":"mechinterp"},{"id":"2505.24539","title":"Localizing Persona Representations in LLMs","created":"2025-05-30","cited_by_count":3,"area":null},{"id":"2510.04500","title":"Expand Neurons, Not Parameters","created":"2025-10-06","cited_by_count":3,"area":null},{"id":"2511.06516","title":"You Had One Job: Per-Task Quantization Using LLMs' Hidden Representations","created":"2025-11-09","cited_by_count":3,"area":null},{"id":"2604.06005","title":"Disentangling MLP Neuron Weights in Vocabulary Space","created":"2026-04-07","cited_by_count":3,"area":"mechinterp"},{"id":"2605.14694","title":"The Rate-Distortion-Polysemanticity Tradeoff in SAEs","created":"2026-05-14","cited_by_count":3,"area":"mechinterp"},{"id":"2310.04519","title":"SPADE: Sparsity-Guided Debugging for Deep Neural Networks","created":"2023-10-06","cited_by_count":2,"area":null},{"id":"2410.20161","title":"Causal Abstraction in Model Interpretability: A Compact Survey","created":"2024-10-26","cited_by_count":2,"area":null},{"id":"2502.20130","title":"QPM: Discrete Optimization for Globally Interpretable Image Classification","created":"2025-02-27","cited_by_count":2,"area":null},{"id":"2505.23117","title":"Decom-Renorm-Merge: Model Merging on the Right Space Improves Multitasking","created":"2025-05-29","cited_by_count":2,"area":null},{"id":"2509.00928","title":"Superposition in Graph Neural Networks","created":"2025-08-31","cited_by_count":2,"area":null},{"id":"2510.17256","title":"Explainability of Large Language Models: Opportunities and Challenges toward Generating Trustworthy Explanations","created":"2025-10-20","cited_by_count":2,"area":"mechinterp"},{"id":"2510.23671","title":"Sparsity and Superposition in Mixture of Experts","created":"2025-10-26","cited_by_count":2,"area":"moe"},{"id":"2603.10771","title":"Word Recovery in Large Language Models Enables Character-Level Tokenization Robustness","created":"2026-03-11","cited_by_count":2,"area":null},{"id":"2604.04037","title":"Geometric Limits of Knowledge Distillation: A Minimum-Width Theorem via Superposition Theory","created":"2026-04-05","cited_by_count":2,"area":"distill"},{"id":"2608.27540","title":"Towards a mathematical theory of superposition","created":"2026-08-27","cited_by_count":2,"area":null},{"id":"2410.06235","title":"Parameter Choice and Neuro-Symbolic Approaches for Deep Domain-Invariant Learning","created":"2024-10-08","cited_by_count":1,"area":null},{"id":"2507.09816","title":"Compressed Computation: Dense Circuits in a Toy Model of the Universal-AND Problem","created":"2025-07-13","cited_by_count":1,"area":null},{"id":"2603.00270","title":"LLMs Remember First, Forget Last: Dual-Process Interference in Large Language Models","created":"2026-02-27","cited_by_count":1,"area":null},{"id":"2605.02234","title":"Bucketing the Good Apples: A Method for Diagnosing and Improving Causal Abstraction","created":"2026-05-04","cited_by_count":1,"area":null},{"id":"2605.12770","title":"WriteSAE: Sparse Autoencoders for Recurrent State","created":"2026-05-12","cited_by_count":1,"area":"ssm"},{"id":"2605.12809","title":"Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces","created":"2026-05-12","cited_by_count":1,"area":null},{"id":"2605.16339","title":"Preference Instability in Reward Models: Detection and Mitigation via Sparse Autoencoders","created":"2026-05-07","cited_by_count":1,"area":"alignment"},{"id":"2606.03990","title":"Neuron Populations Exhibit Divergent Selectivity with Scale","created":"2026-06-02","cited_by_count":1,"area":null},{"id":"2406.18926","title":"Fine-tuned network relies on generic representation to solve unseen cognitive task","created":"2024-06-27","cited_by_count":0,"area":null},{"id":"2502.15090","title":"ExpertLens: Activation steering features are highly interpretable","created":"2025-02-20","cited_by_count":0,"area":"mechinterp"},{"id":"2506.21624","title":"DCN^2: Interplay of Implicit Collision Weights and Explicit Cross Layers for Large-Scale Recommendation","created":"2025-06-24","cited_by_count":0,"area":null},{"id":"2508.06592","title":"Towards Integrated Alignment","created":"2025-08-08","cited_by_count":0,"area":"alignment"},{"id":"2601.07411","title":"SCALPEL: Selective Capability Ablation via Low-rank Parameter Editing for Large Language Model Interpretability Analysis","created":"2026-01-12","cited_by_count":0,"area":null},{"id":"2605.05223","title":"Structural Instability of Feature Composition","created":"2026-04-18","cited_by_count":0,"area":"mechinterp"},{"id":"2605.23040","title":"Steered Generation via Gradient-Based Optimization on Sparse Query Features","created":"2026-05-21","cited_by_count":0,"area":"mechinterp"},{"id":"2605.31304","title":"Interpretability Without Tradeoffs: Disentangling Polysemanticity At Equal Predictive Performance","created":"2026-05-29","cited_by_count":0,"area":null},{"id":"2606.18897","title":"SAERec: Constructing Fine-grained Interpretable Intents Priors via Sparse Autoencoders for Recommendation","created":"2026-06-17","cited_by_count":0,"area":"mechinterp"},{"id":"2606.27440","title":"PairSAE: Mechanistic Interpretability from Pair Representations in Protein Co-Folding","created":"2026-06-25","cited_by_count":0,"area":"mechinterp"},{"id":"2609.14861","title":"Semantic Fibers and Cross-Gram Interference: A Calculus of Safety Drift in Overcomplete Representations","created":"2026-09-14","cited_by_count":0,"area":"safety"},{"id":"2609.32966","title":"Self-Confirming Superposition Traps in Reinforcement Learning","created":"2026-09-26","cited_by_count":0,"area":"deeprl"},{"id":"2609.35351","title":"Interference Beyond Geometry in Concept Extraction","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2609.36455","title":"Emergent phases of superposition: from partial to full representation","created":"2026-09-29","cited_by_count":0,"area":null},{"id":"2610.04905","title":"ScopeSAE: Model-Scope Feature Discovery with Interpretable Layer Selection","created":"2026-10-04","cited_by_count":0,"area":"mechinterp"}],"next_cursor":null}