{"total":146,"rows":[{"id":"2502.02013","title":"Layer by Layer: Uncovering Hidden Representations in Language Models","created":"2025-02-04","cited_by_count":311,"area":null},{"id":"2409.05907","title":"Programming Refusal with Conditional Activation Steering","created":"2024-09-06","cited_by_count":206,"area":"mechinterp"},{"id":"2503.03321","title":"See What You Are Told: Visual Attention Sink in Large Multimodal Models","created":"2025-03-05","cited_by_count":132,"area":"mllm"},{"id":"2411.11296","title":"Steering Language Model Refusal with Sparse Autoencoders","created":"2024-11-18","cited_by_count":95,"area":"mechinterp"},{"id":"2503.06287","title":"Your Large Vision-Language Model Only Needs A Few Attention Heads For Visual Grounding","created":"2025-03-08","cited_by_count":83,"area":"mllm"},{"id":"2505.20063","title":"SAEs Are Good for Steering -- If You Select the Right Features","created":"2025-05-26","cited_by_count":78,"area":"mechinterp"},{"id":"2607.15495","title":"Verbalizable Representations Form a Global Workspace in Language Models","created":"2026-07-16","cited_by_count":75,"area":null},{"id":"2503.13423","title":"SuperBPE: Space Travel for Language Models","created":"2025-03-17","cited_by_count":63,"area":null},{"id":"2502.05795","title":"The Curse of Depth in Large Language Models","created":"2025-02-09","cited_by_count":62,"area":null},{"id":"2501.06346","title":"Large Language Models Share Representations of Latent Grammatical Concepts Across Typologically Diverse Languages","created":"2025-01-10","cited_by_count":59,"area":null},{"id":"2407.09298","title":"Transformer Layers as Painters","created":"2024-07-12","cited_by_count":54,"area":null},{"id":"2510.06477","title":"Attention Sinks and Compression Valleys in LLMs are Two Sides of the Same Coin","created":"2025-10-07","cited_by_count":52,"area":null},{"id":"2505.13898","title":"Do Language Models Use Their Depth Efficiently?","created":"2025-05-20","cited_by_count":50,"area":null},{"id":"2410.05864","title":"From Tokens to Words: On the Inner Lexicon of LLMs","created":"2024-10-08","cited_by_count":48,"area":null},{"id":"2601.04480","title":"When Models Manipulate Manifolds: The Geometry of a Counting Task","created":"2026-01-08","cited_by_count":46,"area":null},{"id":"2410.14670","title":"Decomposing The Dark Matter of Sparse Autoencoders","created":"2024-10-18","cited_by_count":45,"area":"mechinterp"},{"id":"2410.08201","title":"Efficient Dictionary Learning with Switch Sparse Autoencoders","created":"2024-10-10","cited_by_count":39,"area":"mechinterp"},{"id":"2412.13795","title":"Mix-LN: Unleashing the Power of Deeper Layers by Combining Pre-LN and Post-LN","created":"2024-12-18","cited_by_count":37,"area":null},{"id":"2412.09563","title":"Does Representation Matter? Exploring Intermediate Layers in Large Language Models","created":"2024-12-12","cited_by_count":34,"area":null},{"id":"2604.11791","title":"A Mechanistic Analysis of Looped Reasoning Language Models","created":"2026-04-13","cited_by_count":33,"area":"lr"},{"id":"2509.12886","title":"The LLM Already Knows: Estimating LLM-Perceived Question Difficulty via Hidden Representations","created":"2025-09-16","cited_by_count":29,"area":null},{"id":"2409.05448","title":"Representational Analysis of Binding in Language Models","created":"2024-09-09","cited_by_count":28,"area":null},{"id":"2503.08200","title":"Route Sparse Autoencoder to Interpret Large Language Models","created":"2025-03-11","cited_by_count":26,"area":"mechinterp"},{"id":"2409.03621","title":"Attend First, Consolidate Later: On the Importance of Attention in Different LLM Layers","created":"2024-09-05","cited_by_count":25,"area":null},{"id":"2408.01416","title":"The Quest for the Right Mediator: Surveying Mechanistic Interpretability Through the Lens of Causal Mediation Analysis","created":"2024-08-02","cited_by_count":24,"area":"mechinterp"},{"id":"2504.03022","title":"The Dual-Route Model of Induction","created":"2025-04-03","cited_by_count":24,"area":null},{"id":"2409.04185","title":"Residual Stream Analysis with Multi-Layer SAEs","created":"2024-09-06","cited_by_count":19,"area":"mechinterp"},{"id":"2609.01343","title":"SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers","created":"2026-09-01","cited_by_count":18,"area":"moe"},{"id":"2605.23872","title":"Training-Free Looped Transformers","created":"2026-05-22","cited_by_count":15,"area":"moe"},{"id":"2602.20031","title":"Latent Introspection: Models Can Detect Prior Concept Injections","created":"2026-02-23","cited_by_count":14,"area":null},{"id":"2504.01928","title":"Is the Reversal Curse a Binding Problem? Uncovering Limitations of Transformers from a Basic Generalization Failure","created":"2025-04-02","cited_by_count":13,"area":null},{"id":"2604.09780","title":"The Myth of Expert Specialization in MoEs: Why Routing Reflects Geometry, Not Necessarily Domain Expertise","created":"2026-04-10","cited_by_count":13,"area":"moe"},{"id":"2410.11042","title":"Persistent Topological Features in Large Language Models","created":"2024-10-14","cited_by_count":12,"area":null},{"id":"2504.08775","title":"Layers at Similar Depths Generate Similar Activations Across LLM Architectures","created":"2025-04-03","cited_by_count":12,"area":null},{"id":"2505.18658","title":"Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metrics","created":"2025-05-24","cited_by_count":12,"area":null},{"id":"2506.21103","title":"Learning to Skip the Middle Layers of Transformers","created":"2025-06-26","cited_by_count":10,"area":"moe"},{"id":"2605.06510","title":"Is One Layer Enough? Understanding Inference Dynamics in Tabular Foundation Models","created":"2026-05-07","cited_by_count":10,"area":null},{"id":"2506.22638","title":"Layer Importance for Mathematical Reasoning is Forged in Pre-Training and Invariant after Post-Training","created":"2025-06-27","cited_by_count":9,"area":"lr"},{"id":"2509.17588","title":"Interpreting Attention Heads for Image-to-Text Information Flow in Large Vision-Language Models","created":"2025-09-22","cited_by_count":9,"area":"mllm"},{"id":"2602.05970","title":"Inverse Depth Scaling From Most Layers Being Similar","created":"2026-02-05","cited_by_count":9,"area":null},{"id":"2501.19406","title":"Low-Rank Adapting Models for Sparse Autoencoders","created":"2025-01-31","cited_by_count":7,"area":"mechinterp"},{"id":"2502.02790","title":"Leveraging the true depth of LLMs","created":"2025-02-05","cited_by_count":7,"area":null},{"id":"2506.00653","title":"Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models","created":"2025-05-31","cited_by_count":7,"area":null},{"id":"2511.21581","title":"Learning When to Stop: Adaptive Latent Reasoning via Reinforcement Learning","created":"2025-11-26","cited_by_count":7,"area":"lr"},{"id":"2603.01502","title":"Geometric Alignment without Functional Equivalence: A Layer-wise Analysis of the Speech-Text Modality Gap","created":"2026-03-02","cited_by_count":7,"area":null},{"id":"2412.18239","title":"OMG-HD: A High-Resolution AI Weather Model for End-to-End Forecasts from Observations","created":"2024-12-24","cited_by_count":6,"area":null},{"id":"2505.20133","title":"Token Distillation: Attention-aware Input Embeddings For New Tokens","created":"2025-05-26","cited_by_count":6,"area":null},{"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":"2506.12913","title":"Jailbreak Transferability Emerges from Shared Representations","created":"2025-06-15","cited_by_count":6,"area":"safety"},{"id":"2509.05291","title":"Crosscoding Through Time: Tracking Emergence & Consolidation Of Linguistic Representations Throughout LLM Pretraining","created":"2025-09-05","cited_by_count":6,"area":null},{"id":"2509.23782","title":"Bridging the Knowledge-Prediction Gap in LLMs on Multiple-Choice Questions","created":"2025-09-28","cited_by_count":6,"area":null},{"id":"2512.08819","title":"Do Depth-Grown Models Overcome the Curse of Depth? An In-Depth Analysis","created":"2025-12-09","cited_by_count":6,"area":null},{"id":"2603.03308","title":"Old Habits Die Hard: How Conversational History Geometrically Traps LLMs","created":"2026-02-08","cited_by_count":6,"area":null},{"id":"2409.17113","title":"Characterizing stable regions in the residual stream of LLMs","created":"2024-09-25","cited_by_count":5,"area":null},{"id":"2501.15754","title":"Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference","created":"2025-01-27","cited_by_count":5,"area":null},{"id":"2508.14067","title":"Punctuation and Predicates in Language Models","created":"2025-08-11","cited_by_count":5,"area":null},{"id":"2603.05235","title":"Reclaiming Lost Text Layers for Source-Free Cross-Domain Few-Shot Learning","created":"2026-03-05","cited_by_count":5,"area":null},{"id":"2604.19052","title":"Cell-Based Representation of Relational Binding in Language Models","created":"2026-04-21","cited_by_count":5,"area":null},{"id":"2410.08255","title":"Investigating Representation Universality: Case Study on Genealogical Representations","created":"2024-10-10","cited_by_count":4,"area":null},{"id":"2502.04030","title":"Fine, I'll Merge It Myself: A Multi-Fidelity Framework for Automated Model Merging","created":"2025-02-06","cited_by_count":4,"area":"lr"},{"id":"2506.10641","title":"Spelling-out is not Straightforward: LLMs' Capability of Tokenization from Token to Characters","created":"2025-06-12","cited_by_count":4,"area":null},{"id":"2511.18162","title":"Vector Arithmetic in Concept and Token Subspaces","created":"2025-11-22","cited_by_count":4,"area":null},{"id":"2511.21594","title":"Visualizing LLM Latent Space Geometry Through Dimensionality Reduction","created":"2025-11-26","cited_by_count":4,"area":null},{"id":"2603.15389","title":"When Does Sparsity Mitigate the Curse of Depth in LLMs","created":"2026-03-16","cited_by_count":4,"area":null},{"id":"2604.12151","title":"Distinct mechanisms underlying in-context learning in transformers","created":"2026-04-14","cited_by_count":4,"area":null},{"id":"2607.19139","title":"Text Template Tokens Are Implicit Semantic Registers in Diffusion Transformers","created":"2026-07-21","cited_by_count":4,"area":"diffusion"},{"id":"2608.10837","title":"TACTICL: Task-Aware Compression of Tabular ICL Models","created":"2026-08-11","cited_by_count":4,"area":null},{"id":"2504.14107","title":"Signatures of human-like processing in Transformer forward passes","created":"2025-04-18","cited_by_count":3,"area":null},{"id":"2601.03779","title":"Tracing the complexity profiles of different linguistic phenomena through the intrinsic dimension of LLM representations","created":"2026-01-07","cited_by_count":3,"area":null},{"id":"2601.08017","title":"Representations of Text and Images Align From Layer One","created":"2026-01-12","cited_by_count":3,"area":"mllm"},{"id":"2601.18302","title":"Suppressing Final Layer Hidden State Jumps in Transformer Pretraining","created":"2026-01-26","cited_by_count":3,"area":null},{"id":"2602.01997","title":"On the Limits of Layer Pruning for Generative Reasoning in Large Language Models","created":"2026-02-02","cited_by_count":3,"area":"lr"},{"id":"2602.14759","title":"Inner Loop Inference for Pretrained Transformers: Unlocking Latent Capabilities Without Training","created":"2026-02-16","cited_by_count":3,"area":null},{"id":"2603.06922","title":"NerVE: Nonlinear Eigenspectrum Dynamics in LLM Feed-Forward Networks","created":"2026-03-06","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":"2605.23033","title":"Uncovering the Latent Potential of Deep Intermediate Representations","created":"2026-05-21","cited_by_count":3,"area":null},{"id":"2606.20560","title":"How Transparent is DiffusionGemma?","created":"2026-06-18","cited_by_count":3,"area":"lr"},{"id":"2609.07876","title":"LLM Layers Immediately Correct Each Other","created":"2026-09-07","cited_by_count":3,"area":"mechinterp"},{"id":"2407.04513","title":"LayerShuffle: Enhancing Robustness in Vision Transformers by Randomizing Layer Execution Order","created":"2024-07-05","cited_by_count":2,"area":null},{"id":"2502.11646","title":"Hyper-SET: Designing Transformers via Hyperspherical Energy Minimization","created":"2025-02-17","cited_by_count":2,"area":null},{"id":"2504.13763","title":"Decoding Vision Transformers: the Diffusion Steering Lens","created":"2025-04-18","cited_by_count":2,"area":"mechinterp"},{"id":"2507.12950","title":"Insights into a radiology-specialised multimodal large language model with sparse autoencoders","created":"2025-07-17","cited_by_count":2,"area":"mllm"},{"id":"2509.04655","title":"Polysemantic Dropout: Conformal OOD Detection for Specialized LLMs","created":"2025-09-04","cited_by_count":2,"area":null},{"id":"2601.21621","title":"Similarity of Processing Steps in Vision Model Representations","created":"2026-01-29","cited_by_count":2,"area":null},{"id":"2602.03134","title":"SwiftVLM: Efficient Vision-Language Model Inference via Cross-Layer Token Bypass","created":"2026-02-03","cited_by_count":2,"area":"mllm"},{"id":"2602.06020","title":"Two Stages of Folding: Convergent Mechanisms in AI Protein Folding Trunks","created":"2026-02-05","cited_by_count":2,"area":null},{"id":"2602.18960","title":"Modularity is the Bedrock of Natural and Artificial Intelligence","created":"2026-02-21","cited_by_count":2,"area":null},{"id":"2602.21750","title":"From Words to Amino Acids: Does the Curse of Depth Persist?","created":"2026-02-25","cited_by_count":2,"area":null},{"id":"2604.16656","title":"Defragmenting Language Models: An Interpretability-based Approach for Vocabulary Expansion","created":"2026-04-17","cited_by_count":2,"area":null},{"id":"2606.23670","title":"Tapered Language Models","created":"2026-06-22","cited_by_count":2,"area":null},{"id":"2606.26472","title":"Epiphany-Aware KV Cache Eviction Without the Attention Matrix","created":"2026-06-25","cited_by_count":2,"area":"lr"},{"id":"2607.00627","title":"AGI Maze as a Benchmark Framework for World-Modeling Agents","created":"2026-07-01","cited_by_count":2,"area":"wm"},{"id":"2411.07071","title":"Universal Response and Emergence of Induction in LLMs","created":"2024-11-11","cited_by_count":1,"area":null},{"id":"2503.18908","title":"FFN Fusion: Rethinking Sequential Computation in Large Language Models","created":"2025-03-24","cited_by_count":1,"area":"lr"},{"id":"2503.23924","title":"Model Hemorrhage and the Robustness Limits of Large Language Models","created":"2025-03-31","cited_by_count":1,"area":null},{"id":"2505.16178","title":"Understanding Fact Recall in Language Models: Why Two-Stage Training Encourages Memorization but Mixed Training Teaches Knowledge","created":"2025-05-22","cited_by_count":1,"area":null},{"id":"2505.17936","title":"Understanding Gated Neurons in Transformers from Their Input-Output Functionality","created":"2025-05-23","cited_by_count":1,"area":null},{"id":"2506.02973","title":"Expanding before Inferring: Enhancing Factuality in Large Language Models through Premature Layers Interpolation","created":"2025-06-03","cited_by_count":1,"area":null},{"id":"2601.03464","title":"Prompting Underestimates LLM Capability for Time Series Classification","created":"2026-01-06","cited_by_count":1,"area":"mechinterp"}],"next_cursor":"100"}