{"type":"post","id":"jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632","slug":"jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632","title":"JailBound: FOL-guided jailbreak evaluation accepted at NeurIPS 2026","description":"JailBound is a jailbreak-evaluation framework that generates prompts from a structured threat taxonomy and searches for behavior-sensitive model states.","retrieval_nugget":"JailBound is a jailbreak-evaluation framework that generates prompts from a structured threat taxonomy and searches for behavior-sensitive model states. Its authors report tests across 46 language models, where one search branch exceeded a zero-order baseline on attack success rate, giving safety teams a more systematic way to extend testing beyond fixed prompt lists.","published_at":"2026-11-06","updated_at":"2026-11-06","record_date":"2026-11-06","date_kind":"published_at","topics":["prompts"],"entities":["JailBound","NeurIPS","S-Eval"],"related_patterns":["ai-engineering-evaluations-are-splitting-into-task-specific"],"novelty_label":"new_fact","what_changed":"JailBound adds a new FOL-guided jailbreak evaluation framework, reporting results across 46 LLMs and improved attack success over a ZOL-only baseline.","prior_refs":[],"source_url":"https://arxiv.org/abs/2611.07530","source_urls":["https://arxiv.org/abs/2611.07530"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632","status":"published","visuals":[],"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632/","markdown":"https://newruntime.com/posts/jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632.md","json":"https://newruntime.com/posts/jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632.json"},"markdown_url":"https://newruntime.com/posts/jailbound-fol-guided-jailbreak-evaluation-accepted-at-neurips-e4d632.md","next_reads":[{"type":"topic","path":"/topics/prompts/","url":"https://newruntime.com/topics/prompts/","title":"Prompts","reason":"Explore the prompts topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/prompts.md","json_url":"https://newruntime.com/topics/prompts.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/sources/","url":"https://newruntime.com/sources/","title":"Sources","reason":"See the original sources this archive cites.","media_type":"text/html"}]}