8:23Kunal Singh - Do third-party frontier AI evaluations matter? [MATS Symposium | Summer 2026]MATS Research2251 mo ago
11:44Tom Kimpson - Can Power Draw Survive Covert Compute? [MATS Symposium | Summer 2026]MATS Research1911 mo ago
7:17Andrew Jun Lee - LLMs Develop Belief State Geometry In-Context [MATS Symposium | Summer 2026]MATS Research8921 mo ago
10:43Ortega-Fernández, Kowalczyk - Detecting Steganographic Weight Exfil [MATS Symposium | Summer 2026]MATS Research801 mo ago
14:02Rishi Jha - The Road to Hell Is Paved with Persistent Agents [MATS Symposium | Summer 2026]MATS Research1871 mo ago
11:30John Kitaoka & Thao Pham - Red-Teaming Multi-Agent Attacks [MATS Symposium | Summer 2026]MATS Research1541 mo ago
10:35Reworr & Sree Sharvesh - Stress-testing loss-of-control threats [MATS Symposium | Summer 2026]MATS Research1241 mo ago
6:33Ihor Protsenko - Towards scaling laws of emergent deceptive alignment [MATS Symposium | Summer 2026]MATS Research1171 mo ago
8:51Michael Ofengenden - Time-Awareness in AI Agents [MATS Symposium | Summer 2026]MATS Research1391 mo ago
8:45Wen Xing - How good are NLA and J-Space for monitoring? [MATS Symposium | Summer 2026]MATS Research4681 mo ago
11:56Joan Velja, Lennie Wells - Building an empirical Science of AI Debate [MATS Symposium | Winter 2026]MATS Research2766 mo ago
9:05Jo Jiao, Boyd Kane - Limited-parameter finetuning [MATS Symposium | Winter 2026]MATS Research4006 mo ago
9:54Arya Jakkli - Eliciting Hidden Behaviors in Large Language Models [MATS Symposium | Winter 2026]MATS Research4916 mo ago
11:02Aria Wong - Mitigating Reward Hacking with RL Training Interventions [MATS Symposium | Winter 2026]MATS Research3146 mo ago
8:40Simon Schrodi - Science of Entangled Capabilities [MATS Symposium | Winter 2026]MATS Research3526 mo ago
11:25T.Buckworth, R.Moreno, A.Panda, Y.Volkova - Control Evals for CMFT [MATS Symposium | Winter 2026]MATS Research2246 mo ago
14:56Nataliia Babina - Measuring LLM Reasoning Consistency [MATS Symposium | Winter 2026]MATS Research2406 mo ago
9:54Yueh-Han “John” Chen - CoT Controllability [MATS Symposium | Summer 2025]MATS Research1011 yr ago
9:29Joschka Braun, Damon Falck, Eyon Jang - Exploration Hacking [MATS Symposium | Summer 2025]MATS Research1031 yr ago
10:24Rich Barton-Cooper, Simon Storf - Black-Box Scheming Monitors [MATS Symposium | Summer 2025]MATS Research751 yr ago
14:44Daniel Reuter - Technical Transparency for AI Labs [MATS Symposium | Summer 2025]MATS Research2371 yr ago
8:09Bruce W. Lee - Training AI Agents to Self-Report Misbehavior [MATS Symposium | Summer 2025]MATS Research861 yr ago
11:26Uzay Macar, Paul Bogdan - Chain-of-Thought Interpretability [MATS Symposium | Summer 2025]MATS Research3331 yr ago
10:21Felix Choussat - How Falling AI Costs Threaten Global Security [MATS Symposium | Summer 2025]MATS Research1401 yr ago
13:37Manfred Diaz - Reinforcement Learning Overoptimization [MATS Symposium | Summer 2025]MATS Research611 yr ago