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
7,673 results for “keyword:"ai"” in ICLR 2027. 1,836 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Chat Templates Are a Hidden Instruction Channel: Inference-Time Backdoors in LLM-Based Systems▲3842.67
[PDF]Revisiting the 16× Pretraining Compute Gap between Diffusion and Autoregressive Language Models▼4.3k2.00
[PDF]Learning to Verify: A Counterfactual-Guided Dual-Loop Agent Framework for Multimodal Hallucination Mitigation▲3.7k2.50
[PDF]OSFP4: Joint Optimization of Diagonal Smoothing and Block Scales for NVFP4 Quantization▲6.1k2.00
[PDF]SysRepair-Bench: Measuring Whether Agents Can Fix a Vulnerability Without Breaking the Service▼4.5k2.67
[PDF]Hard, Not Rare: Human Item Difficulty, Not Concept Frequency, Orders Model Errors Across Professional Examinations▼5.9k2.33
[PDF]Complex Phase Retrieval via Conditional Quantiles: Finite-Sample Recovery without Noise Moments▲133.33
[PDF]Control Protocols for Mitigating Covert Malicious Fine-Tuning Attacks from Internally Deployed Agents▼4.3k2.00
[PDF]Predicting Where Recall Fails: The Fact Matters More Than the Entity in Black-Box Language Models▼1361.33
Withdrawn or desk rejected (1,836)
[PDF]Everything in Moderation: Interior Per-Domain Coverage Optima and Gap-Preserving Fixed-Budget Alignment in Multi-Domain Mid-TrainingDesk rejected
[PDF]When Is a Policy Violation a Model Failure? Policy Observability and Failure Attribution in Agent-Safety EvaluationDesk rejected