Auriga agent harness closes 68.3% of the planning-execution gap in zero-shot robot manipulation
Chengcheng Wang · hf · 2026-11-10
StellarEdge AI released the Auriga technical report, a task-agnostic agent harness for zero-shot embodied manipulation. On RoboDojo, GPT-6 Astra's success rate jumps from 22.5% to 76.2% when arm execution is removed, exposing a large planning-execution gap. Auriga combines explicit task beliefs, calibrated 3D measurements, dual-arm control and execution monitoring to close 68.3% of that gap without target-task trials, achieving state-of-the-art results on RoboDojo and RoboCasa365.
More from AGI Musings
- Demis Hassabis on what schools should look like in the AGI era, echoing Kieran Egan's education theory — round · 2026-11-10
- Blogger predicts AI capabilities will plateau as lab valuations stall and open-source catches up — isnit0 · 2026-11-10
- Kenneth Stanley's team releases open-endedness research agenda: blog + 4 preprints — kenneth0stanley · 2026-11-10
- Gabriel (littmath) on why compute asymmetry is distorting scientific credit and publishing incentives — littmath · 2026-11-10
- One interactive timeline of every tech and pharma leader's disease-cure prediction — jungofthewon · 2026-11-10
- AGI risk skepticism fades as serious thinkers now take it seriously, observer says — danfaggella · 2026-11-10