residual stream

How Context Shapes Truth: Geometric Transformations of Statement-level Truth Representations in LLMs ACL 2026 Sycophancy Hides Linearly in the Attention Heads EACL 2026 Where Paths Split: Localized, Calibrated Control of Moral Reasoning in Large Language Models ACL 2026 Breaking Bad Tokens: Detoxification of LLMs Using Sparse Autoencoders EMNLP 2025 Calibration Across Layers: Understanding Calibration Evolution in LLMs EMNLP 2025 Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders ACL 2025 From Mechanistic Interpretability to Mechanistic Biology: Training, Evaluating, and Interpreting Sparse Autoencoders on Protein Language Models ICML 2025 How Do Transformers Learn Variable Binding in Symbolic Programs? ICML 2025 HyperDAS: Towards Automating Mechanistic Interpretability with Hypernetworks ICLR 2025 ICR Probe: Tracking Hidden State Dynamics for Reliable Hallucination Detection in LLMs ACL 2025 Improving Reasoning Performance in Large Language Models via Representation Engineering ICLR 2025 Measuring In-Context Computation Complexity via Hidden State Prediction ICML 2025 Residual Matrix Transformers: Scaling the Size of the Residual Stream ICML 2025 Residual Stream Analysis with Multi-Layer SAEs ICLR 2025 Revealing the Deceptiveness of Knowledge Editing: A Mechanistic Analysis of Superficial Editing ACL 2025 Robust LLM safeguarding via refusal feature adversarial training ICLR 2025 Scaling Effects on Latent Representation Edits in GPT Models (Student Abstract) AAAI 2025 Sparse Autoencoders Reveal Temporal Difference Learning in Large Language Models ICLR 2025 The Reasoning-Memorization Interplay in Language Models Is Mediated by a Single Direction ACL 2025 Why Do Some Inputs Break Low-Bit LLM Quantization? EMNLP 2025 An Adversarial Example for Direct Logit Attribution: Memory Management in GELU-4L EMNLP 2024 Neurons in Large Language Models: Dead, N-gram, Positional ACL 2024 On Large Language Models’ Hallucination with Regard to Known Facts NAACL 2024 On the Similarity of Circuits across Languages: a Case Study on the Subject-verb Agreement Task EMNLP 2024 Refusal in Language Models Is Mediated by a Single Direction NeurIPS 2024