CoACT: Action-Preserving Observation Compression for Coding Agents Paper • 2607.02911 • Published 26 days ago
Grammar-Constrained Decoding Can Jailbreak LLMs into Generating Malicious Code Paper • 2606.11817 • Published Jun 10 • 18
M^3Eval: Multi-Modal Memory Evaluation through Cognitively-Grounded Video Tasks Paper • 2606.05008 • Published Jun 3 • 26