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
104 results for “keyword:"safety alignment"” in ICLR 2027. 24 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]TutelaBench: Evaluating the Safety of LLM Planners for Care Robots Among Simulated Humans▼4.3k2.00
[PDF]HiRoute: Hierarchical Routed Prompt Tuning for Safety Alignment of Large Language Models▼1.2k1.67
[PDF]Improving Safety Alignment and Health Safe Completion via Query-Specific Safety Rubrics in OPSD▲13k2.33
[PDF]SafetyTransfer: Distilling Harmful and Harmless Discrimination for Selective Refusal in Small Language Models▼4.3k2.00
[PDF]SafeSchatten: Schatten-Norm-Constrained Steepest Descent in Activation Space for Safe LLM Fine-Tuning▲3.0k2.33
[PDF]Benign Tasks, Unsafe Orders: The Local-to-Global Safety Gap in Continual LLM Fine-Tuning▼5.8k2.33
[PDF]When Do Model Internals Help? Exploring the Role of Representation Engineering in LLM Safety▼5.8k2.33
[PDF]DACPO: Drift-Adaptive Constrained Policy Optimization for Cue-Free Safety Recovery in Large Reasoning Models▲3952.67
[PDF]Reference-KL Hijacking: Reference Poisoning Attacks in Distributed LLM Reinforcement Learning Frameworks▼4.3k2.00
[PDF]Directing Safety Signals to the Appropriate Neurons: Multilingual Safety Alignment through Selective Neuron Masking▲9.2k2.67
[PDF]DeShortcut-Align: Decoupling Spurious Shortcuts for Robust Safety Alignment in Large Reasoning Models▲5103.00
[PDF]Benign Safety Is Not a Robustness Certificate: Evidence of a Structural Evaluation Gap Across Attack Conditions▲11.00
[PDF]Risk Under Pressure: Compute-Aware Evaluation of Adversarial Robustness in Language Models▲3.0k2.33
[PDF]Seed-Level Instability in LLM Safety Alignment: Regime Shifts in Large-Margin Preference Optimization▼4.3k2.00
[PDF]Safety Is Discernment, Not Refusal: Teaching Large Reasoning Models to Judge Before They Act▲3952.67
[PDF]Beyond ASR: Evaluating Safety Alignment of Multimodal Large Language Models Across Modalities and Low-Resource Languages▼1.2k1.67
[PDF]Learning to Stay Safe: Adaptive Regularization Against Safety Degradation during Fine-Tuning▼1361.33
[PDF]One Calibration, Many Requests: Amortized Black-Box Multimodal Jailbreaking via Cross-Model Refusal Coordinates▲14k3.00
[PDF]Empathetic yet Guarded: Dynamic Personalized Safety Alignment in Multi-Turn Conversations▲3.0k2.33
Withdrawn or desk rejected (24)
[PDF]Why Benign Post-Training Induces Safety Gate Collapse: Changes in Harmful Activation Geometry Relative to the Inherited Refusal GateWithdrawn1.67
[PDF]Erased But Not Gone: How Model Merging Weakens Concept Erasure in Diffusion ModelsWithdrawn1.33