A Tsinghua and University of Pennsylvania team has open-sourced OpenMise, a spatial scene editing toolkit positioned as a bridge between agents and 3D scenes. It splits hard video generation into two simpler stages, building an editable 3D scene first and rendering video second, fixing detail drift such as misplaced chess pieces or clipping and cutting repeated re-rolls.
Companiesslug: tsinghua-university
Tsinghua University
清华大学
Leading Chinese university and co-author of the TCSExam benchmark paper.
Curated Intelligence
新智元 (微信公众号)T2·Open Source
74 pts
Rohan Paul (@rohanpaul_ai)T2·Models
78 pts
Tsinghua University and Moonshot AI released TCSExam, a 100-task expert-reviewed benchmark spanning Lean theorem proving, solver engineering, record-setting constructions, and 30 open conjectures, with every submission verified by the Lean kernel or a task checker. GPT-6 Astra fully proved 70.0% of theorems without supplied lemmas, versus 38.3% for Kimi K3.1 Pro and at most 11.7% for six other agents, yet the same model solved none of the 30 open TCS conjectures; the few accepted conjecture submissions merely adapted counterexamples and proofs OpenAI had already published.