List your company

Robotic task planning and code generation on Advanced

90Success RateCorrective GSCE
-3.620.74569.3Apr 13, 2026

Evaluation results

MethodLinks
Corrective GSCE2026.049097.8
Corrective GSCE2026.0481.796.3
GSCE2026.047591.5
Ro-SLMtraining tasks=Complex Task Only2026.047583.8
GSCE2026.047594.8
Ro-SLMvariant=without GRPO Optimization2026.047586.4
Ro-SLM2026.047587
Ro-SLMGRPO optimization=none2026.047083.1
Ro-SLM2026.047083.7
Ro-SLMvariant=without Augmentation2026.047081.3
Ro-SLMvariant=Complex Task Only2026.047083
Ro-SLMaugmentation=none2026.045574.8
Ro-SLMvariant=with Comments2026.045575.2
Ro-SLMcomments=retained2026.045076.2
Llama-3.1-8B2026.0459
Ro-SLMtraining tasks=Simple Task Only2026.0401.9
Llama-3.1-8B2026.0408.4
Ro-SLMvariant=Simple Task Only2026.0403.6