NeurIPS 2026 reviews are up: reviews, scores, rebuttals and meta-reviews for 7,877 papers. See the highest rated
acceptodds: the papers of ICLR 2027
1,261 results for “keyword:"interpretability"” in ICLR 2027. 264 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Answer Candidates Reshape Language Model Decisions: Layer-wise Dynamics and Causal State Interventions▲6.1k2.00
[PDF]Localization Does Not Guarantee Control: Gate Failure and Partial Activation Steering in a Causal Reasoning Model▼1.2k1.67
[PDF]TARGEN: A STRUCTURAL SCORE BEYOND MOLECULAR EVIDENCE FOR CANCER TARGET GENE PREDICTION▲3.0k2.33
[PDF]Training Selects Among Behaviorally Equivalent Computations in In-Context Dictionary Learning▲5.4k3.00
[PDF]Nuisance-Induced Confounding in LLM Output Sensitivity Analysis: A Semantic Normalization Framework▼4.3k2.00
[PDF]From Post-Training Plasticity to Function-Targeted Adaptation in Large Language Models▼4.3k2.00
[PDF]Linguistic Cartography: Metric Spatial and Hierarchical Platial Representations in Language Models▲5103.00
[PDF]A Polanyi-Like Paradox in Large Language Models: The Gap Between Spatial Representation and Decision▼5.8k2.33
[PDF]Right Answer, Wrong Reasoning: Localizing Chain-of-Thought Restoration Errors in the Residual Stream▲3952.67
[PDF]Representing and Steering a Hypothesized Consciousness Spectrum: Manifold-Guided vs. Linear Directions▼4.3k2.00
[PDF]Weights Read and Write Features: Scalable Parameter Decomposition Grounded in Activation Space▲3952.67
[PDF]The Sink Tax: Auditing Memory Tokens and Assigning Their Roles in Long-Context Language Models▼4.3k2.00
[PDF]Extracting and Composing Function Vectors for Analogical Reasoning over Latent Relations▲9.2k2.67
Withdrawn or desk rejected (264)
[PDF]A Window Is Not a Door: Mapping the Limits and Conditions of Knowledge Editing in Language ModelsDesk rejected2.33