MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head Paper • 2601.07832 • Published 9 days ago • 47
Watching, Reasoning, and Searching: A Video Deep Research Benchmark on Open Web for Agentic Video Reasoning Paper • 2601.06943 • Published 10 days ago • 202
AT^2PO: Agentic Turn-based Policy Optimization via Tree Search Paper • 2601.04767 • Published 13 days ago • 26
kailinjiang/llava_1.5_13b_covariance_matrices_from_onevision_pre_64_seed_rank233_new222 Updated 22 days ago
kailinjiang/llava_1.5_13b_covariance_matrices_from_onevision_pre_64_seed_rank233_new Updated 22 days ago
TongSIM: A General Platform for Simulating Intelligent Machines Paper • 2512.20206 • Published 29 days ago • 28
KORE Collection KORE uses knowledge-oriented control as its pivot to synergistically optimize the balance between knowledge adaptation and retention at different stag • 30 items • Updated Dec 4, 2025 • 1