{"total":85,"rows":[{"id":"2405.00208","title":"A Primer on the Inner Workings of Transformer-based Language Models","created":"2024-04-30","cited_by_count":85,"area":null},{"id":"2409.03752","title":"Attention Heads of Large Language Models: A Survey","created":"2024-09-05","cited_by_count":67,"area":null},{"id":"2505.13763","title":"Language Models Are Capable of Metacognitive Monitoring and Control of Their Internal Activations","created":"2025-05-19","cited_by_count":41,"area":null},{"id":"2503.02854","title":"(How) Do Language Models Track State?","created":"2025-03-04","cited_by_count":34,"area":null},{"id":"2603.17839","title":"How do LLMs Compute Verbal Confidence","created":"2026-03-18","cited_by_count":26,"area":null},{"id":"2409.09951","title":"Optimal ablation for interpretability","created":"2024-09-16","cited_by_count":25,"area":null},{"id":"2407.04690","title":"Missed Causes and Ambiguous Effects: Counterfactuals Pose Challenges for Interpreting Neural Networks","created":"2024-07-05","cited_by_count":23,"area":null},{"id":"2412.04703","title":"Transformers Struggle to Learn to Search","created":"2024-12-06","cited_by_count":19,"area":null},{"id":"2405.17345","title":"Exploring and steering the moral compass of Large Language Models","created":"2024-05-27","cited_by_count":16,"area":null},{"id":"2412.12276","title":"Emergence and Effectiveness of Task Vectors in In-Context Learning: An Encoder Decoder Perspective","created":"2024-12-16","cited_by_count":9,"area":null},{"id":"2602.16698","title":"Causality is Key for Interpretability Claims to Generalise","created":"2026-02-18","cited_by_count":8,"area":null},{"id":"2505.08546","title":"Are We Paying Attention to Her? Investigating Gender Disambiguation and Attention in Machine Translation","created":"2025-05-13","cited_by_count":7,"area":null},{"id":"2512.11508","title":"On Geometric Understanding and Learned Priors in Feed-forward 3D Reconstruction Models","created":"2025-12-12","cited_by_count":7,"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":"2502.12055","title":"Designing Role Vectors to Improve LLM Inference Behaviour","created":"2025-02-17","cited_by_count":6,"area":null},{"id":"2506.13060","title":"Rethinking Explainability in the Era of Multimodal AI","created":"2025-06-16","cited_by_count":6,"area":null},{"id":"2506.16702","title":"Large Language Models as Psychological Simulators: A Methodological Guide","created":"2025-06-20","cited_by_count":6,"area":null},{"id":"2606.00831","title":"Subliminal Learning is a LoRA Artifact","created":"2026-05-30","cited_by_count":6,"area":null},{"id":"2512.07383","title":"LogicCBMs: Logic-Enhanced Concept-Based Learning","created":"2025-12-08","cited_by_count":5,"area":null},{"id":"2601.22851","title":"When Meanings Meet: Investigating the Emergence and Quality of Shared Concept Spaces during Multilingual Language Model Training","created":"2026-01-30","cited_by_count":5,"area":null},{"id":"2604.19884","title":"From Signal Degradation to Computation Collapse: Uncovering the Two Failure Modes of LLM Quantization","created":"2026-04-21","cited_by_count":5,"area":null},{"id":"2507.00322","title":"Failure by Interference: Language Models Make Balanced Parentheses Errors When Faulty Mechanisms Overshadow Sound Ones","created":"2025-06-30","cited_by_count":4,"area":null},{"id":"2601.16407","title":"Jacobian Scopes: token-level causal attributions in LLMs","created":"2026-01-23","cited_by_count":4,"area":null},{"id":"2502.11806","title":"Exploring the Translation Mechanism of Large Language Models","created":"2025-02-17","cited_by_count":3,"area":null},{"id":"2505.13514","title":"Induction Head Toxicity Mechanistically Explains Repetition Curse in Large Language Models","created":"2025-05-17","cited_by_count":3,"area":null},{"id":"2506.01723","title":"Tug-of-war between idioms' figurative and literal interpretations in LLMs","created":"2025-06-02","cited_by_count":3,"area":null},{"id":"2506.20746","title":"Dynamic Weight Grafting: Localizing Finetuned Factual Knowledge in Transformers","created":"2025-06-25","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":"2602.00767","title":"BLOCK-EM: Preventing Emergent Misalignment via Latent Blocking","created":"2026-01-31","cited_by_count":3,"area":null},{"id":"2604.22271","title":"How LLMs Detect and Correct Their Own Errors: The Role of Internal Confidence Signals","created":"2026-04-24","cited_by_count":3,"area":null},{"id":"2503.15551","title":"Efficient but Vulnerable: Benchmarking and Defending LLM Batch Prompting Attack","created":"2025-03-18","cited_by_count":2,"area":null},{"id":"2509.23095","title":"Causally-Enhanced Reinforcement Policy Optimization","created":"2025-09-27","cited_by_count":2,"area":null},{"id":"2510.11218","title":"The Curious Case of Factual (Mis)Alignment between LLMs' Short- and Long-Form Answers","created":"2025-10-13","cited_by_count":2,"area":null},{"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":"2604.15400","title":"Hallucination as Trajectory Commitment: Causal Evidence for Asymmetric Attractor Dynamics in Transformer Generation","created":"2026-04-16","cited_by_count":2,"area":null},{"id":"2605.09314","title":"How LLMs Are Persuaded: A Few Attention Heads, Rerouted","created":"2026-05-10","cited_by_count":2,"area":null},{"id":"2608.14894","title":"Can Neural Networks Learn by Experimenting on Themselves? Self-Interventional Learning from Functional Consequences to Predictive Self-Knowledge","created":"2026-08-14","cited_by_count":2,"area":null},{"id":"2608.26511","title":"Sycophancy Suppression Can Impair Rational Updating: Anti-Sycophancy Should Preserve the Ability to Update","created":"2026-08-27","cited_by_count":2,"area":null},{"id":"2410.01334","title":"Skill Path: Unveiling Language Skills from Circuit Graphs","created":"2024-10-02","cited_by_count":1,"area":null},{"id":"2410.03447","title":"How Language Models Prioritize Contextual Grammatical Cues?","created":"2024-10-04","cited_by_count":1,"area":null},{"id":"2511.10694","title":"Where does an LLM begin computing an instruction?","created":"2025-11-12","cited_by_count":1,"area":null},{"id":"2511.13371","title":"Cognitive Maps in Language Models: A Mechanistic Analysis of Spatial Planning","created":"2025-11-17","cited_by_count":1,"area":null},{"id":"2601.03570","title":"How Do Large Language Models Learn Concepts During Continual Pre-Training?","created":"2026-01-07","cited_by_count":1,"area":null},{"id":"2604.23452","title":"From Edges to Depth: Probing the Spatial Hierarchy in Vision Transformers","created":"2026-04-25","cited_by_count":1,"area":null},{"id":"2605.09195","title":"The Geometry of Forgetting: Temporal Knowledge Drift as an Independent Axis in LLM Representations","created":"2026-05-09","cited_by_count":1,"area":null},{"id":"2605.24624","title":"Vision-Language Binding in In-Context Image Generation","created":"2026-05-23","cited_by_count":1,"area":null},{"id":"2606.25010","title":"Emergent Capabilities Arise Randomly from Learning Sparse Attention Patterns","created":"2026-06-23","cited_by_count":1,"area":null},{"id":"2607.06648","title":"Final Checkpoints Are Not Enough: Analyzing Latent Reasoning Faithfulness Along Training Trajectories","created":"2026-07-07","cited_by_count":1,"area":null},{"id":"2607.13047","title":"Targeted Recovery of Weight-Space Mechanisms From Neural Networks","created":"2026-06-19","cited_by_count":1,"area":null},{"id":"2608.01816","title":"Divergent large language model predictions from convergent representations in ambiguous word pairs","created":"2026-08-03","cited_by_count":1,"area":null},{"id":"2608.04183","title":"Test, then Route: How Language Models Execute In-Context Conditional Rules Across Models and Languages","created":"2026-08-04","cited_by_count":1,"area":null},{"id":"2609.07183","title":"CircuitLens: Reasoning Circuits as Data Selection Signals for Reinforcement Learning with Verifiable Rewards","created":"2026-09-07","cited_by_count":1,"area":null},{"id":"2501.00885","title":"Representation in large language models","created":"2025-01-01","cited_by_count":0,"area":null},{"id":"2507.10106","title":"BlueGlass: A Framework for Composite AI Safety","created":"2025-07-14","cited_by_count":0,"area":null},{"id":"2510.14336","title":"DARTS-GT: Differentiable Architecture Search for Graph Transformers with Quantifiable Instance-Specific Interpretability Analysis","created":"2025-10-16","cited_by_count":0,"area":null},{"id":"2602.14370","title":"Competition for attention predicts good-to-bad tipping in AI","created":"2026-02-16","cited_by_count":0,"area":null},{"id":"2602.19109","title":"Post-Routing Arithmetic in Llama-3: Last-Token Result Writing and Rotation-Structured Digit Directions","created":"2026-02-22","cited_by_count":0,"area":null},{"id":"2603.12423","title":"Interpreting Negation in GPT-2: Layer- and Head-Level Causal Analysis","created":"2026-03-12","cited_by_count":0,"area":null},{"id":"2604.14433","title":"Zero-Ablation Overstates Register Content Dependence in DINO Vision Transformers","created":"2026-04-15","cited_by_count":0,"area":null},{"id":"2605.07284","title":"Same Targets, Different Computation: How Post-Training Divides Work Across Model Layers","created":"2026-05-08","cited_by_count":0,"area":null},{"id":"2605.27033","title":"Tracing Computation Density in LLMs","created":"2026-05-26","cited_by_count":0,"area":null},{"id":"2606.08292","title":"Necessary, Decodable and Reversible, Yet Not Transferable: A Stress Test for Attention-Head Role Claims","created":"2026-06-06","cited_by_count":0,"area":null},{"id":"2606.21078","title":"Validation-Gated Causal Interventions for Interpreting High-Stakes Large Language Model Behavior: A Case Study in Suicidality Detection","created":"2026-06-19","cited_by_count":0,"area":null},{"id":"2606.30096","title":"Information Dynamics of Language Communication","created":"2026-06-29","cited_by_count":0,"area":null},{"id":"2607.04846","title":"Pretraining Curricula Enable Selective Fine-tuning","created":"2026-07-06","cited_by_count":0,"area":null},{"id":"2607.20995","title":"Where Animacy Lives in Large Language Models: Tracing the Circuits of the Animacy Concept","created":"2026-07-23","cited_by_count":0,"area":null},{"id":"2607.21491","title":"What, Where, and How: Disentangling the Roles of Task, Language, and Model in Code Model Representations","created":"2026-07-23","cited_by_count":0,"area":null},{"id":"2607.28672","title":"LAWFUL: Law-Aligned Witness for Faithful Use of Latents","created":"2026-07-26","cited_by_count":0,"area":null},{"id":"2608.04347","title":"Introspecting Alignment Shifts Beyond Behaviors Implanted Through Fine-Tuning","created":"2026-08-05","cited_by_count":0,"area":null},{"id":"2608.15772","title":"Broken Symmetry in LLM Refusal: Answer Release Is More Local Than Refusal Restoration","created":"2026-08-16","cited_by_count":0,"area":null},{"id":"2608.18768","title":"Readable, Faithful, Used: Three Dissociable Properties of Demographic Identity in a Language Model","created":"2026-08-19","cited_by_count":0,"area":null},{"id":"2609.00515","title":"The Interlingua Hypothesis: LLMs Translate via a Latent Task-agnostic Feature Space","created":"2026-09-01","cited_by_count":0,"area":null},{"id":"2609.01081","title":"StateSwap: Probing Support-Elimination Hidden States in Multiple-Choice Questions","created":"2026-09-01","cited_by_count":0,"area":null},{"id":"2609.03322","title":"How Perturbations Propagate: A Multi-Level Analysis of Robustness in Large Language Models","created":"2026-09-03","cited_by_count":0,"area":null},{"id":"2609.10993","title":"Distribution-aware Language Neuron Identification in Multilingual Large Language Models","created":"2026-09-10","cited_by_count":0,"area":null},{"id":"2609.17537","title":"Relation Before Entity: Deferred Commitment in Language Model Factual Recall","created":"2026-07-13","cited_by_count":0,"area":null},{"id":"2609.22224","title":"From Trait Vectors to Circuits: Tracing Refusal and Sycophancy Through Language Models","created":"2026-09-02","cited_by_count":0,"area":null},{"id":"2609.35108","title":"A mechanistic study of language model introspection","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2609.36063","title":"Understanding Decision-Making Mechanisms in Neural Routing Solvers","created":"2026-09-28","cited_by_count":0,"area":null},{"id":"2610.02173","title":"Every Ablation Is a Dose: Counterweights and the Semblance of Self-Repair","created":"2026-10-01","cited_by_count":0,"area":null},{"id":"2610.07270","title":"Does the Model Use the Feature? Separating Steering from Mechanism in LLMs","created":"2026-10-05","cited_by_count":0,"area":null},{"id":"2610.07798","title":"Thin Evidence, Thick Priors: How Language Models Substitute Identity for Missing Financial Facts","created":"2026-10-06","cited_by_count":0,"area":null},{"id":"2610.13794","title":"Not Just an Evaluation: A Case Study of Discourse Particle Sensitivity in Language Models","created":"2026-10-09","cited_by_count":0,"area":null},{"id":"2610.17093","title":"SFT Leaves a Reproducible Signature in Activation Geometry: Calibrating Post-Training Interpretability Against Run-to-Run Noise","created":"2026-10-12","cited_by_count":0,"area":null},{"id":"2610.19113","title":"From Copying to Composition: How Language Models Develop In-Context Learning","created":"2026-10-13","cited_by_count":0,"area":null}],"next_cursor":null}