SenseTime open-sources SenseNova-RoboRSI harness and a 27B embodied model
SenseTime has open-sourced the SenseNova-RoboRSI evaluation harness and a 27B embodied model. The release follows a study of GPT-6-Astra across 12 simulated benchmarks and 8 real robots. The figures are SenseTime's own.
SenseTime has open-sourced the SenseNova-RoboRSI evaluation harness together with a 27B embodied model, according to Pandaily. The release follows a study of GPT-6-Astra that covered 12 simulated benchmarks and 8 real robots. All figures are SenseTime's own.
What was released
- SenseNova-RoboRSI: an evaluation harness for embodied AI, now open source.
- 27B embodied model: a 27-billion-parameter model positioned as a robot brain, also open source.
The study behind it
SenseTime studied GPT-6-Astra across 12 simulated benchmarks and 8 real robots. Those results are company-reported and have not been independently verified.
What this means
The release pairs a test harness with a model, so outside developers can run embodied evaluations and work with a 27B robot model directly. The study's scope, 12 simulated benchmarks plus 8 physical robots, covers both simulation and real hardware. Because the figures come from SenseTime, they should be treated as company claims until others reproduce them with the open-sourced harness.
Related Articles
Shengshu Technology livestreams full Motubrain 2.0 Physical AI training for about half a month
On October 21, Shengshu Technology's world model team began a public livestream of the full Motubrain 2.0 training process, planned to run about half a month. The embodied model is due to be open-sourced, and the stream covers pretraining, post-training, agent task execution and recursive self-improvement (RSI) exploration.
Microsoft's Probabilistic Bias Correction improves ECMWF forecasts for 98% of precipitation targets, runs on CPUs
Microsoft Research has published Probabilistic Bias Correction (PBC) in Nature and released it as open-source software. The researchers report it improved a leading ECMWF physics-based model for 98% of precipitation targets and 92% of temperature targets at 2–6 week lead times, and that it runs on CPUs.
Anthropic Says Claude Proved Security Properties for 239 Entry Points Across 123 Libraries
Anthropic published research on October 22, 2026, reporting that Claude produced machine-checked formal proofs of security properties at the RISC-V assembly level. The proofs cover 239 fuzz-test entry points across 123 open-source libraries, in a collaboration with Theorem.
Microsoft's ThinkingBox-Training lifts Qwen3.8-27B fully reliable task completions from 38 to 107 of 507
Microsoft and Hugging Face released ThinkingBox-Training, which turns executable database-state checks into RL rewards. After 50 GRPO steps, Qwen3.8-27B's pass@1 rose from 51.70% to 60.78%, and tasks passed on all 20 attempts rose from 38 to 107, according to the authors.
Comments
Loading...