Benchmark Ia
Topic archive • 4 matches
AI-generated: summaries written by AI from the linked sources. How we use AI
2026-11-12
Technology
Google DeepMind released Gemini 4 Hydrogen, its third new model in six weeks, which introduces an xhigh reasoning level. According to Artificial Analysis' Intelligence Index, Hydrogen nearly ties Claude Sonnet 5.5 and GPT-6.1 Sol, at a third of Sonnet's cost but roughly 2.5x that of Sol. The model is rolling out in Google Search and inside Google Antigravity alongside a new Orchestrator option for long-horizon tasks.
AI Models • Techmeme
Permalink
2026-11-14
Technology
Runway released Praxis-1, a robotics model with open weights and source code under the Apache 2.0 license. The model transfers knowledge from video training to robot control and leads the World Action Model category on Nvidia's RoboLab-120 benchmark, though it cannot yet plan complex action sequences on its own.
Robotics • The Decoder
Permalink
2026-11-15
Technology
UniPat AI's SPU benchmark asks AI agents to rank the outcomes of 4,207 real wet-lab experiments, and none of the 13 configurations tested scored above 0.5. The best configuration averaged 0.486 out of 1, highlighting how far agents remain from reliably predicting experimental results.
Research • Pandaily
Permalink