Goal images used as control targets in robotic manipulation are not best chosen by general image‑editing quality. The authors show that large proprietary editors, though top on editing‑quality leaderboards, produce poorer control targets than smaller fine‑tuned open models, especially on long‑horizon tasks. They introduce SGE‑Goal, a spatially‑grounded editing framework that predicts edit‑region masks before synthesis, enabling open base models to become the strongest goal generators on CALVIN and nearly as strong on SIMPLER‑Bridge, with superior performance on chained tasks.
arXiv:2607. 11643v1 Announce Type: cross Abstract: Recent foundation image and video generation models offer strong generalization and controllability, but their direct application to embodied scenarios is limited by requirements for multi-view consistency, geometric coherence, and robot embodiment constraints.
arXiv:2610.10409v1 Announce Type: cross Abstract: General-purpose agents increasingly write code, use tools, and complete complex digital tasks, raising the question of how far these capabilities car...
arXiv:2606. 10025v1 Announce Type: cross Abstract: We present GHOST, a framework for learning visuomotor manipulation policies that generalize beyond the training distribution.
The paper introduces Interaction‑Aligned Pruning (IAprune), a training‑free method for visual token pruning in embodied manipulation tasks. IAprune jointly decides per‑frame budget and token selection, using semantic‑motion spatial agreement to choose between conservative and aggressive coverage, and applies geometric residual correction to focus on under‑represented boundaries. Experiments on four policies, three simulation benchmarks, and a real‑robot platform show that IAprune matches unpruned performance on LIBERO while achieving up to 1.54× speed‑up and 1.48× acceleration on a real robot.
The paper introduces a closed‑loop framework to evaluate frontier AI models on the LIBERO‑10 robot manipulation benchmark, allowing models to control a robot via visual inputs, proprioception, and tool calls without task‑specific fine‑tuning. Six models (GPT‑6 Astra, GPT‑6.1 Sol, GPT‑6 Luna, Claude Opus 5, Claude Opus 5.5, Gemini‑4 Argon) were tested across 18 configurations and 1,800 episodes, with GPT‑6 Astra achieving the highest success rate of 44%. Results show that higher reasoning effort generally improves performance, yet benefits vary by model and task, and many models struggle with precise spatial arrangements and complex multi‑step interactions.
arXiv:2606. 28385v1 Announce Type: cross Abstract: Recent advances in robot world models enable synthetic video generation for embodied prediction and planning.
Recent advances in frontier AI models have demonstrated remarkable capabilities in visual understanding, reasoning, planning, and tool use, raising the possibility of using general-purpose AI models f...
arXiv:2606. 17846v1 Announce Type: cross Abstract: Foundation models in language and multimodality achieve strong generalization by aligning heterogeneous data under a unified formulation and training at scale.
Recent image generators have demonstrated impressive photorealism and instruction-following capabilities in single-image generation and editing. However, constrained by their architectures, they cannot achieve interleaved generation (text-image sequence), which has crucial applications in visual narratives, guidance, and embodied manipulation.
AFFORD2ACT investigates whether robot manipulation policies can rely on a very compact visual representation instead of dense images. Using an affordance model to select up to 15 object keypoints and 4 end‑effector keypoints, the policy receives only 19 2D points (38 coordinates). In six real‑robot tasks, this sparse keypoint input outperforms RGB, RGB‑D, and dense‑affordance baselines, maintains performance with fewer demonstrations, and achieves 82% success on held‑out data, though it depends on accurate affordance localization and keypoint tracking.
AntiGrounding is a visual action-selection framework that turns short robot trajectories into both executable motion plans and rendered prompts for vision‑language model evaluation. After filtering for feasibility, each trajectory is scored on safety, task alignment, efficiency, and physical plausibility using structured multi‑view visual question answering, and the best trajectories are refined and validated by a digital twin before real‑world execution. In eight real‑world manipulation tasks, the system achieved a 71.25% success rate with a single GPT‑6 Astra evaluator, outperforming baseline methods.
Recent breakthroughs in instruction-based image editing have captured significant attention, as models are now capable of handling real-world editing demands with the practicality required by everyday users. However, editing models trained primarily for single-turn edits often break down in multi-turn editing--the natural interactive setting where a user iteratively refines an image based on the model's own previous outputs.