RoboJev-27B is an open‑weight vision‑language model that predicts whether a robot has successfully completed a manipulation task by analyzing a task instruction, ten frames of the attempt, and optional auxiliary camera frames. Trained with a rank‑32 LoRA adapter on 52,598 episodes, it achieves 95.2% balanced accuracy and 0.993 AUROC on held‑out episodes of the same robot model, and maintains strong performance on unseen labs, robots, and community‑collected data. The model runs in 0.26 seconds per episode on an H100 GPU and is released with inference code and a large, transparent evaluation dataset for reproducible benchmarking.
arXiv:2608. 11739v1 Announce Type: cross Abstract: The prevailing recipe for Vision-Language-Action (VLA) models couples a pretrained VLM with a separately trained flow-matching action expert.
IndustrialVLA-Bench evaluates six released vision‑language‑action (VLA) and world‑action (WAM) robot policy models using a unified reporting schema. It measures clean capability on LIBERO, non‑language robustness on LIBERO‑Plus, instruction sensitivity on LIBERO‑Para, and observed execution cost, aggregating results across three random seeds. The benchmark also reports inference latency, peak memory, runtime mode, and evidence status for each system, enabling traceable comparisons without claiming one paradigm is universally superior.
arXiv:2609.24170v1 Announce Type: new Abstract: Embodied AI systems are often organized into System 1 and System 2. System 1 is typically a pretrained policy that generates actions at high frequency,...
arXiv:2606. 26443v1 Announce Type: cross Abstract: A robot working alongside people must reason about what they have done, in what order, and with what intent.
arXiv:2607. 25487v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models translate natural-language commands into robot action sequences, but leading systems on the LIBERO-Plus robustness benchmark use three- to seven-billion-parameter backbones whose memory demands can exceed embedded robotic budgets.
The paper introduces a perception interface that separates vision from language in vision‑language models. A frozen perception stack detects objects, a deterministic semantic serializer converts the perceived state into text, and a standard text‑only large language model (LLM) answers questions. Experiments on a campus‑robot benchmark show that this serialized interface outperforms a zero‑shot VLM of the same language‑model size, especially as the language model shrinks, and that the advantage persists under paraphrase and different supervision regimes.
RoboSPA is a large-scale robotic manipulation dataset and benchmark designed to evaluate Vision‑Language‑Action models on fine‑grained spatial reasoning and long‑horizon procedural planning. It contains 10 task categories, 56 base tasks, and 280 variants across five difficulty levels, with 527K trajectories collected from multiple embodiments and scenes. The benchmark introduces diagnostic metrics beyond binary success, revealing that current VLA models struggle with complex spatial relations, precise execution, and memory‑intensive planning.
REKI is a 2.4B vision‑language‑action model that achieves a new balance between performance and size, outperforming only larger models on several benchmarks. It combines a vision‑language model with a flow‑matching action expert and incorporates a lightweight memory of recent frames via register tokens, which reduces pose oscillations and improves success in both simulation and real‑world robot tests. The authors provide full code, data curation, training, inference, evaluation, deployment scripts, and all pretrained and fine‑tuned weights on GitHub.
arXiv:2608.21407v1 Announce Type: cross Abstract: Vision-language-action (VLA) models face a crucial tradeoff between their task success rate and the policy-call frequency. Executing a single action...
REKI is a 2.4B vision‑language‑action model that achieves a new Pareto frontier between performance and size on several robotics benchmarks. It combines a vision‑language model with a flow‑matching action expert and adds a lightweight memory of recent frames via register tokens, reducing pose oscillations and improving success rates. The authors provide full code, data curation, training, inference, evaluation, deployment scripts, and all pretrained and finetuned weights on GitHub.
INSPECT is a system that learns how a robot should choose its camera view during assembly inspection by observing a smart‑glasses assistant that answers part queries and guides the user. It uses techniques such as Presence‑Invariant TwinSwap for object evidence calibration, claim‑indexed supervision to separate evidence needs from camera changes, and object‑centered calibration to adapt view preferences to robot poses. In experiments on gearbox assemblies and angle‑grinder recordings, INSPECT outperforms other non‑oracle policies, improving view utility and decision accuracy.
VLCP (Vision Language Control Policy) is a training‑free robot manipulation approach that keeps a vision‑language model (VLM) frozen and uses it to generate short Python control functions. Unlike traditional methods that retry a fixed policy, VLCP rewrites the control code every K steps based on multi‑view RGB, proprioceptive state, and state delta, allowing failures to be corrected within the same episode. In a 57‑task MuJoCo/RoboVerse benchmark, VLCP achieves 35.1% pooled success versus 3.5% for a single‑query baseline, with a 27.3% within‑episode recovery rate on failed grasps and efficient token usage.