DeepMind unveils Gemini Robotics 2.5 and ASIMOV-jailbreak benchmark
Apollo 2 robots kitted parts in Mercedes factories at 96% success; the open jailbreak set has 1,500 prompts.
Stories per day · 30 days
9 RSS
Relatedrobotics25vla8world-models8benchmark5vision-language-action5vlm4benchmarks3diffusion3libero3self-improvement3simulation3erqa2mujoco2multimodal2physical-reasoning2planning2reinforcement-learning2robot-learning2
112 stories, newest first · one card per story
Stories tagged #embodied-ai
Embodied First Exam tests coding agents on 1,156 robot tasks
GPT-6 Astra in Codex, strongest of eight agents, succeeds on 34% of 456 core tasks.
LingBot-Touch learns shared tactile tokens across sensor types
Authors say the tokens plug into diffusion policies and VLA models, including pi0.5 and LingBot-VLA.
Computer-Use Agents Function as Embodied Generalists
Computer-use agents win multiple ECCV competitions by treating embodied tasks as software interactions.
A2-MEAS embodied framework reports BEHAVIOR and real-robot scores
Authors report 0.6411 Q-score on 100 public tasks and 57.92% success on 12 real-robot tasks versus 26.67% for pi0.5.
VLA-Tug probe improves failure detection in VLAs
VLA-Tug uses latent probing to detect and classify failures in VLAs, improving early detection by 30% and accuracy on unseen tasks by 44%.
BEAKER benchmark evaluates embodied MLLMs in chemical labs
BEAKER reveals gaps in action-grounded perception for self-driving chemical laboratories using 1,000 expert-annotated samples.
EviSeek Framework Improves Embodied Question Answering
EviSeek makes evidence demands explicit to boost accuracy on 3,436 ESI-Bench questions.
PyRUA-REAL enables robot exploration and adaptive learning through recursive self-improvement
The PyRUA-REAL framework allows coding agents to retain interaction experience as persistent memory, achieving 91.1% success on eight real-world tasks.
Open-HumanScale Scales Human Video for Robot Learning
Open-HumanScale-50K corpus and HumanWAM models pretrained on 50k hours of egocentric video achieve high success rates on robot benchmarks.
TriPhysBench evaluates physical intelligence coordination in vision-language models
Benchmark shows top vision-language models achieve only 50.38% accuracy on integrated physical tasks.
Auriga harness doubles GPT-6 Astra zero-shot robot success on RoboDojo
The frozen harness posts 57.7% success and stays first under scene randomization while learned policies fall by up to 33.9%.
ID-WM decouples reachability from control in hierarchical world models
ID-WM uses high-level sampling of feasible transitions and low-level inverse dynamics solving to enable scalable long-horizon planning with 38.5% better performance.
REA runtime coordinates slow planners with fast robot policies
On 3,000 LIBERO-Pro episodes it reports 95.9% success, 25% fewer control steps, and 17% less control time than waiting on the brain.
Gym-Anything builds RoboWorld, 25 coding-agent robotics simulations
Frontier models score an average reward of 0.06 and fail every RoboWorld test task; several teaching methods raise performance.
PDRVLA conditions robot actions on physical dynamics without video generation
It reports 99.3% success on LIBERO and 81.5% on SimplerEnv using public data and a frozen world model.
EmbodiedLab benchmarks robots that experiment in order to learn
Across 30 simulated science experiments, even the strongest of six frontier agents succeeds less than half the time.
The Trinity evolves multimodal-agent harnesses from shared capabilities
The authors report mean gains of 34 points on 28 training benchmarks and 26 points on 18 unseen benchmarks.
RynnVLA-Latent pretrains a 4B robot policy on 3D latent actions
A 4B policy pretrained on 745.52M descriptors from 30.1K hours of video posts strong LIBERO, VLABench, and RoboTwin scores.
HuGe pretrains humanoid VLAs on 2,000 hours of egocentric data
The three-stage pipeline reports state-of-the-art simulation results and higher Unitree G1 pick-and-drop success.
SceneBloom generates collision-free indoor layouts for embodied agent training
Navigation policies trained on its scenes beat policies trained on the same number of real scans in held-out scans.
Sensorimotor Alignment proposed as offline proxy for robot policy success
The training-free S^2 metric correlates with closed-loop robot task success and can rank VLM backbones before VLA fine-tuning.
RynnUnify 16B model unifies embodied understanding, planning, and control
Authors report a 9.75-point Rynn-VP gain over Cosmos 3 Nano and 46.7% average success on three real-robot tasks.
New offline metric measures sensorimotor alignment in AI models
Sensorimotor Alignment S2 correlates with robot task success and identifies strong vision-language-action models before expensive rollouts.