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
11 results for “keyword:"security benchmark"” in ICLR 2027. 1 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]DuelBench: A Benchmark for Closed-Loop Attack and Defense Between LLM Agents on Web Services▼5.8k2.33
[PDF]Do Defenses Against LLM Extraction Work Across Attacks? A Lifecycle Benchmark of Black-Box Model Extraction▲14k3.00
[PDF]SysRepair-Bench: Measuring Whether Agents Can Fix a Vulnerability Without Breaking the Service▼4.5k2.67
[PDF]ProtoArena: Can AI Agents Discover and Exploit Real-World Cross-Component Protocol Vulnerabilities?▼4.3k2.00