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
517 results for “keyword:"activation"” in ICLR 2027. 119 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Transformers develop separable posterior and value representations in a simple decision task▼4.3k2.00
[PDF]Identity-Exact Multi-Bias Activations: Breaking Optimization Symmetry Without Changing the Pretrained Function▼1361.33
[PDF]Interventional Commutativity: Effect Preservation and Reference Selection Under Weight-Only Quantisation▼1.2k1.67
[PDF]Gradients for Interventions and Activations for Detection: Targeted Feature Learning in Language Models▼4.3k2.00
[PDF]Multi-Dimensional Featurizers Yield Fine-Grained Steering and Data Shaping in Robot Foundation Models▼5.1k2.60
[PDF]BMIPrune: Class- and Layer-aware One-shot Average Activation-based Pruning for Convolutional Neural Networks via Mutual Information▼4.3k2.00
[PDF]Same Data, Different Target: How the Prediction Target Can Separate Probing from Steering▼1.2k1.67
[PDF]MultiUS : Unsupervised Persona Mixture and Norm-Bounded Latent Steering for User Simulation▼1361.33
[PDF]Prompt-Adapted Transport Control: Contextual Tokenwise LLM Behaviour Steering via Flow Matching▼4.5k2.67
[PDF]Small Transformers Learn Kepler's Equation with a Number Line and an Eccentricity Register▼5.8k2.33
[PDF]Removing the NEEDLE in the Haystack: Backdoor Removal in LLMs via Weight Orthogonalisation▲3952.67
[PDF]Emotion Interpretability in Multimodal Models: Disentangling Behavioral, Representational, and Interventional Evidence▼5.8k2.33
[PDF]Looped Transformers under the Jacobian Lens: Does the Global Workspace Survive Recurrence?▼4.3k2.00
[PDF]Harmonic Variance Embedding: A Positional-Style Representation for Counteracting Depth Dilution in Pre-Norm Transformers▲11.00
[PDF]Evaluating Weight-Level Bias Mitigation Methods Against Behavioral and Activation-Space Bias▼1361.33
[PDF]Learn Here, Move Less Elsewhere: Input-Conditioned Plasticity from Retained-Domain Activation Atlases▼12k1.67
[PDF]StiefelSteer: Invertible Refusal Steering via Low-Dimensional Rotations of the Residual Stream▼1.2k1.67
[PDF]The Hidden Cost of Prompt Formatting: Measuring Representational Sensitivity in Language Models▼1361.33
[PDF]Rethinking Activation Steering for Mathematical Reasoning: What Actually Drives Its Reasoning Gains?▼4.3k2.00
[PDF]BLADE-PDE: Blended-Activation Denoising and LLM-Aided Discovery of Partial Differential Equations from Noisy Data▼4.5k2.25
[PDF]COQUAD: CO-OPTIMIZED QUADRATIC ACTIVATIONS FOR EFFICIENT PRIVATE TRANSFORMER INFERENCE▼4.5k2.67
[PDF]Hallucinations Exhibit Shallower Entropy Valleys: Training-Free Hallucination Detection in LLMs▼1.2k1.67
[PDF]Routing-Guided Single-Expert Interventions Measured Against a Model’s Own Execution Variability▲6.1k2.00
Withdrawn or desk rejected (119)
[PDF]Why Benign Post-Training Induces Safety Gate Collapse: Changes in Harmful Activation Geometry Relative to the Inherited Refusal GateWithdrawn1.67
[PDF]Activation Density in Transformer Feed-Forward Networks Is Set by the Nonlinearity, Not the WidthWithdrawn2.00
[PDF]Concept Subspaces Compute Beyond the Logit Lens: A Weights-Only Test for Locating Representations Upstream of ReadoutWithdrawn