PRTS: A Primitive Reasoning and Tasking System via Contrastive Representations
About
Vision-Language-Action (VLA) models advance robotic control via strong visual-linguistic priors. However, existing VLAs predominantly frame pretraining as supervised behavior cloning, overlooking the fundamental nature of robot learning as a goal-reaching process that requires understanding temporal task progress. We present \textbf{PRTS} (\textbf{P}rimitive \textbf{R}easoning and \textbf{T}asking \textbf{S}ystem), a VLA foundation model that reformulates pretraining through Goal-Conditioned Reinforcement Learning. By treating language instructions as goals and employing contrastive reinforcement learning, PRTS learns a unified embedding space where the inner product of state-action and goal embeddings approximates the log-discounted goal occupancy, the probability of reaching the language-specified goal from the current state-action, quantitatively assessing physical feasibility beyond static semantic matching. PRTS draws this dense goal-reachability supervision directly from offline trajectories without reward annotations, and folds it into the VLM backbone via a role-aware causal mask, incurring negligible overhead over vanilla behavior cloning. This paradigm endows the high-level reasoning system with intrinsic goal reachability awareness, bridging semantic reasoning and temporal task progress, and further benefits goal-conditioned action prediction. Pretrained on 167B tokens of diverse manipulation and embodied-reasoning data, PRTS reaches state-of-the-art performance on LIBERO, LIBERO-Pro, LIBERO-Plus, SimplerEnv, and a real-world suite of 14 complex tasks, with particularly substantial gains on long-horizon, contact-rich, and zero-shot novel-instruction settings, confirming that injecting goal-reachability awareness significantly improves both execution success and long-horizon planning of general-purpose robotic foundation policies.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Robotic Manipulation | LIBERO-Plus | Language Understanding Score89.6 | 249 | |
| Robot Manipulation | SimplerEnv WidowX Visual Matching | Average Success Rate77.1 | 34 | |
| Visuomotor Control | LIBERO | Spatial Score98.8 | 18 | |
| Robotic Manipulation | LIBERO-Pro (test) | Semantic SR97 | 6 | |
| Flip Tennis Tube | RealMan dual-arm platform Real-world | Success Rate90 | 3 | |
| Hand Over | RealMan dual-arm platform Real-world | Success Rate95 | 3 | |
| Office Long Term | RealMan dual-arm platform Real-world | Success Rate95 | 3 | |
| Paper Rubbish | RealMan dual-arm platform Real-world | Success Rate100 | 3 | |
| Pick Shoes | RealMan dual-arm platform Real-world | Success Rate95 | 3 | |
| Serve Tea | RealMan dual-arm platform Real-world | Success Rate95 | 3 |