{"total":92,"rows":[{"id":"2501.17148","title":"AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders","created":"2025-01-28","cited_by_count":218,"area":"mechinterp"},{"id":"2502.16681","title":"Are Sparse Autoencoders Useful? A Case Study in Sparse Probing","created":"2025-02-23","cited_by_count":115,"area":"mechinterp"},{"id":"2411.11296","title":"Steering Language Model Refusal with Sparse Autoencoders","created":"2024-11-18","cited_by_count":95,"area":"mechinterp"},{"id":"2506.19823","title":"Persona Features Control Emergent Misalignment","created":"2025-06-24","cited_by_count":86,"area":null},{"id":"2503.05613","title":"A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Models","created":"2025-03-07","cited_by_count":76,"area":"mechinterp"},{"id":"2501.08319","title":"Enhancing Automated Interpretability with Output-Centric Feature Descriptions","created":"2025-01-14","cited_by_count":41,"area":null},{"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":"2505.16188","title":"SAE-SSV: Supervised Steering in Sparse Representation Spaces for Reliable Control of Language Models","created":"2025-05-22","cited_by_count":29,"area":"mechinterp"},{"id":"2505.20322","title":"Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms","created":"2025-05-23","cited_by_count":24,"area":null},{"id":"2502.17601","title":"Representation Engineering for Large-Language Models: Survey and Research Challenges","created":"2025-02-24","cited_by_count":21,"area":"mechinterp"},{"id":"2412.17626","title":"Tracking the Feature Dynamics in LLM Training: A Mechanistic Study","created":"2024-12-23","cited_by_count":16,"area":"mechinterp"},{"id":"2502.14258","title":"Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information","created":"2025-02-20","cited_by_count":15,"area":null},{"id":"2505.20254","title":"Position: Mechanistic Interpretability Should Prioritize Feature Consistency in SAEs","created":"2025-05-26","cited_by_count":14,"area":"mechinterp"},{"id":"2502.03032","title":"Analyze Feature Flow to Enhance Interpretation and Steering in Language Models","created":"2025-02-05","cited_by_count":13,"area":"mechinterp"},{"id":"2503.14530","title":"SAUCE: Selective Concept Unlearning in Vision-Language Models with Sparse Autoencoders","created":"2025-03-16","cited_by_count":13,"area":"mllm"},{"id":"2505.16004","title":"Evaluating Adversarial Robustness of Concept Representations in Sparse Autoencoders","created":"2025-05-21","cited_by_count":12,"area":"mechinterp"},{"id":"2507.13410","title":"Causal Language Control in Multilingual Transformers via Sparse Feature Steering","created":"2025-07-17","cited_by_count":12,"area":"mechinterp"},{"id":"2508.12535","title":"CorrSteer: Generation-Time LLM Steering via Correlated Sparse Autoencoder Features","created":"2025-08-18","cited_by_count":12,"area":"mechinterp"},{"id":"2506.05451","title":"Interpretation Meets Safety: A Survey on Interpretation Methods and Tools for Improving LLM Safety","created":"2025-06-05","cited_by_count":10,"area":"safety"},{"id":"2510.03659","title":"Does higher interpretability imply better utility? A Pairwise Analysis on Sparse Autoencoders","created":"2025-10-04","cited_by_count":10,"area":"mechinterp"},{"id":"2603.24543","title":"Analysing the Safety Pitfalls of Steering Vectors","created":"2026-03-25","cited_by_count":10,"area":"safety"},{"id":"2503.06040","title":"Mitigating Memorization in LLMs using Activation Steering","created":"2025-03-08","cited_by_count":8,"area":"mechinterp"},{"id":"2504.15133","title":"EasyEdit2: An Easy-to-use Steering Framework for Editing Large Language Models","created":"2025-04-21","cited_by_count":7,"area":"mechinterp"},{"id":"2505.24360","title":"Interpreting Large Text-to-Image Diffusion Models with Dictionary Learning","created":"2025-05-30","cited_by_count":6,"area":"diffusion"},{"id":"2506.04706","title":"Line of Sight: On Linear Representations in VLLMs","created":"2025-06-05","cited_by_count":6,"area":"mechinterp"},{"id":"2602.06256","title":"Steering Safely or Off a Cliff? Rethinking Specificity and Robustness in Inference-Time Interventions","created":"2026-02-05","cited_by_count":6,"area":"lr"},{"id":"2602.07025","title":"The Geometry of Representational Failures in Vision Language Models","created":"2026-02-02","cited_by_count":6,"area":"mllm"},{"id":"2603.28744","title":"Stop Probing, Start Coding: Why Linear Probes and Sparse Autoencoders Fail at Compositional Generalisation","created":"2026-03-30","cited_by_count":6,"area":"mechinterp"},{"id":"2502.06809","title":"Neurons Speak in Ranges: Breaking Free from Discrete Neuronal Attribution","created":"2025-02-04","cited_by_count":5,"area":null},{"id":"2509.12934","title":"The Anatomy of Alignment: Decomposing Preference Optimization by Steering Sparse Features","created":"2025-09-16","cited_by_count":5,"area":"alignment"},{"id":"2605.01844","title":"The Cylindrical Representation Hypothesis for Language Model Steering","created":"2026-05-03","cited_by_count":5,"area":null},{"id":"2606.03002","title":"Perplexity Can Miss SAE Feature Damage Under Quantization","created":"2026-06-02","cited_by_count":5,"area":"mechinterp"},{"id":"2505.08080","title":"Beyond Input Activations: Identifying Influential Latents by Gradient Sparse Autoencoders","created":"2025-05-12","cited_by_count":4,"area":"mechinterp"},{"id":"2510.01246","title":"A Comparative Analysis of Sparse Autoencoder and Activation Difference in Language Model Steering","created":"2025-09-24","cited_by_count":4,"area":"mechinterp"},{"id":"2511.09844","title":"Steering Pretrained Drafters during Speculative Decoding","created":"2025-11-13","cited_by_count":4,"area":"inference"},{"id":"2512.02004","title":"AlignSAE: Concept-Aligned Sparse Autoencoders","created":"2025-12-01","cited_by_count":4,"area":"mechinterp"},{"id":"2512.07667","title":"Depth-Wise Activation Steering for Honest Language Models","created":"2025-12-08","cited_by_count":4,"area":"mechinterp"},{"id":"2602.02224","title":"Spectral Superposition: A Theory of Feature Geometry","created":"2026-02-02","cited_by_count":4,"area":null},{"id":"2605.22658","title":"SegCompass: Exploring Interpretable Alignment with Sparse Autoencoders for Enhanced Reasoning Segmentation","created":"2026-05-21","cited_by_count":4,"area":"lr"},{"id":"2605.23036","title":"Multilingual Steering by Design: Multilingual Sparse Autoencoders and Principled Layer Selection","created":"2026-05-21","cited_by_count":4,"area":"mechinterp"},{"id":"2505.03189","title":"Patterns and Mechanisms of Contrastive Activation Engineering","created":"2025-05-06","cited_by_count":3,"area":"mechinterp"},{"id":"2505.16612","title":"Steering Large Language Models for Machine Translation Personalization","created":"2025-05-22","cited_by_count":3,"area":"mechinterp"},{"id":"2505.16950","title":"Bottlenecked Transformers: Periodic KV Cache Consolidation for Generalised Reasoning","created":"2025-05-22","cited_by_count":3,"area":"lr"},{"id":"2505.17760","title":"But what is your honest answer? Aiding LLM-judges with honest alternatives using steering vectors","created":"2025-05-23","cited_by_count":3,"area":"mechinterp"},{"id":"2507.08967","title":"Self-Improving Model Steering","created":"2025-07-11","cited_by_count":3,"area":"synth"},{"id":"2510.08734","title":"Transmuting prompts into weights","created":"2025-10-09","cited_by_count":3,"area":null},{"id":"2607.25270","title":"Where Steering Signals Come From: Activation Source Selection in Activation Steering","created":"2026-07-28","cited_by_count":3,"area":"mechinterp"},{"id":"2609.07876","title":"LLM Layers Immediately Correct Each Other","created":"2026-09-07","cited_by_count":3,"area":"mechinterp"},{"id":"2511.06048","title":"Visual Exploration of Feature Relationships in Sparse Autoencoders with Curated Concepts","created":"2025-11-08","cited_by_count":2,"area":"mechinterp"},{"id":"2601.08441","title":"YaPO: Learnable Sparse Activation Steering Vectors for Domain Adaptation","created":"2026-01-13","cited_by_count":2,"area":"alignment"},{"id":"2602.10437","title":"Control Reinforcement Learning: Interpretable Token-Level Steering of LLMs via Sparse Autoencoder Features","created":"2026-02-11","cited_by_count":2,"area":"mechinterp"},{"id":"2603.12795","title":"SteerRM: Debiasing Reward Models via Sparse Autoencoders","created":"2026-03-13","cited_by_count":2,"area":"alignment"},{"id":"2603.14278","title":"Activation Surgery: Jailbreaking White-box LLMs without Touching the Prompt","created":"2026-03-15","cited_by_count":2,"area":"safety"},{"id":"2604.10158","title":"Tracing the Thought of a Grandmaster-level Chess-Playing Transformer","created":"2026-04-11","cited_by_count":2,"area":null},{"id":"2606.11205","title":"Dual-Stance Evaluation of Sycophancy: The Structure of Agreement and the Limits of Intervention","created":"2026-04-22","cited_by_count":2,"area":"mechinterp"},{"id":"2502.18644","title":"Steered Generation via Gradient Descent on Sparse Features","created":"2025-02-25","cited_by_count":1,"area":null},{"id":"2506.01770","title":"ReGA: Model-Based Safeguard for LLMs via Representation-Guided Abstraction","created":"2025-06-02","cited_by_count":1,"area":"safety"},{"id":"2506.12576","title":"Calibrating Lightweight Sparse Autoencoder Feature Steering","created":"2025-06-14","cited_by_count":1,"area":"mechinterp"},{"id":"2510.16257","title":"Towards Low-Resource Alignment to Diverse Perspectives with Sparse Feedback","created":"2025-10-17","cited_by_count":1,"area":"alignment"},{"id":"2510.18462","title":"DePass: Unified Feature Attributing by Simple Decomposed Forward Pass","created":"2025-10-21","cited_by_count":1,"area":"mechinterp"},{"id":"2512.15938","title":"SALVE: Sparse Autoencoder-Latent Vector Editing for Mechanistic Control of Neural Networks","created":"2025-12-17","cited_by_count":1,"area":"mechinterp"},{"id":"2605.12813","title":"REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations","created":"2026-05-12","cited_by_count":1,"area":null},{"id":"2606.02528","title":"Auditing Asset-Specific Preferences in Financial Large Language Models: Evidence from Bitcoin Representations and Portfolio Allocation","created":"2026-06-01","cited_by_count":1,"area":null},{"id":"2606.08496","title":"SAEExplainer: Interpreting SAE Features with Activation-Guided Preference Optimization","created":"2026-06-07","cited_by_count":1,"area":"alignment"},{"id":"2607.20596","title":"Are Single-Token Sparse Autoencoder Features Causally Necessary? Layer-Depth and SAE-Family Effects","created":"2026-07-22","cited_by_count":1,"area":"mechinterp"},{"id":"2607.20778","title":"Toward Mechanistic Interpretability of an AI Foundation Model Fine-Tuned for Atmospheric Chemistry","created":"2026-07-22","cited_by_count":1,"area":null},{"id":"2608.28233","title":"REINS: Refusal-Enhanced Inhibitory Steering with Sparse Autoencoder Features","created":"2026-08-28","cited_by_count":1,"area":"safety"},{"id":"2608.28806","title":"Enhancing SAE-based Steering via Neighbor Integrated Feature Selection","created":"2026-08-28","cited_by_count":1,"area":"mechinterp"},{"id":"2506.20083","title":"Bridging Compositional and Distributional Semantics: A Survey on Latent Semantic Geometry via AutoEncoder","created":"2025-06-25","cited_by_count":0,"area":"mechinterp"},{"id":"2602.01716","title":"Mechanistic Indicators of Steering Effectiveness in Large Language Models","created":"2026-02-02","cited_by_count":0,"area":null},{"id":"2602.02280","title":"RACC: Representation-Aware Coverage Criteria for LLM Safety Testing","created":"2026-02-02","cited_by_count":0,"area":"safety"},{"id":"2605.01827","title":"Referring Multiple Regions with Large Multimodal Models via Contextual Latent Steering","created":"2026-05-03","cited_by_count":0,"area":"mllm"},{"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":"2605.28664","title":"Activation Steering for Synthetic Data Generation: The Role of Diversity in Downstream Safety Detection","created":"2026-05-27","cited_by_count":0,"area":"mechinterp"},{"id":"2605.28669","title":"Sense Representations Are Inducible Interfaces","created":"2026-05-27","cited_by_count":0,"area":null},{"id":"2607.13425","title":"Data-Efficient Adaptation of LLMs via Attention Head Reweighting","created":"2026-07-15","cited_by_count":0,"area":null},{"id":"2607.19364","title":"SAE-StatSteer: Statistical Consensus Feature Selection for Optimization-Free Activation Steering of Large Language Models","created":"2026-06-05","cited_by_count":0,"area":"mechinterp"},{"id":"2607.19386","title":"Building Fast, Evaluating Slow: Pipeline Choices Dominate Autointerpretability Score Variance","created":"2026-07-01","cited_by_count":0,"area":"mechinterp"},{"id":"2608.08383","title":"Safety Cost of Steering Vectors Is Separable and Reducible","created":"2026-08-09","cited_by_count":0,"area":"mechinterp"},{"id":"2608.09928","title":"Multimodal Model Diffing for Feature Discovery and Control","created":"2026-08-10","cited_by_count":0,"area":"mllm"},{"id":"2608.13337","title":"Where You Measure Decides What You Measure: Position Selection in Ablation-Based SAE Evaluation","created":"2026-08-13","cited_by_count":0,"area":"mechinterp"},{"id":"2608.30462","title":"Enhancing Low-Resource Language Reasoning via High-Resource Language Feature Transfer","created":"2026-08-31","cited_by_count":0,"area":"lr"},{"id":"2609.10142","title":"Active Adaptation, Not Static Defense: Temporal Dynamics of Preventative Steering in Adversarial Fine-Tuning","created":"2026-09-09","cited_by_count":0,"area":null},{"id":"2609.32355","title":"Gradients for Interventions and Activations for Detection: Targeted Feature Learning in Language Models","created":"2026-09-26","cited_by_count":0,"area":"mechinterp"},{"id":"2609.32519","title":"STR: Supervised Transcoder Replacement for Reducing Steering Side Effects","created":"2026-09-26","cited_by_count":0,"area":"mechinterp"},{"id":"2609.33810","title":"Controlling Speaking Rate in Autoregressive TTS via Activation Steering","created":"2026-09-27","cited_by_count":0,"area":"mechinterp"},{"id":"2609.38718","title":"MetaSteer: Context-Conditioned, nonlinear Steering via Attention-Projection Adaptation","created":"2026-09-30","cited_by_count":0,"area":null},{"id":"2609.39625","title":"D-Scope: Decomposing and Steering Diffusion Transformers with Sparse Autoencoders","created":"2026-09-30","cited_by_count":0,"area":"diffusion"},{"id":"2610.04905","title":"ScopeSAE: Model-Scope Feature Discovery with Interpretable Layer Selection","created":"2026-10-04","cited_by_count":0,"area":"mechinterp"},{"id":"2610.15850","title":"What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions","created":"2026-10-11","cited_by_count":0,"area":"mechinterp"},{"id":"2610.16084","title":"The Sharp Directions Are Against You: Curvature Analysis of Activation Steering","created":"2026-10-12","cited_by_count":0,"area":"mechinterp"},{"id":"2610.17713","title":"Confine to the Task, Anchor the Rest: Concept-Guided Retention Regularizer for Parameter-Efficient Fine-Tuning","created":"2026-10-13","cited_by_count":0,"area":null}],"next_cursor":null}