SimWorld (@simworld_ai) header
SimWorld (@simworld_ai) profile picture

Open-source initiative for open-ended realistic simulation of autonomous agents in physical and social worlds

Joined January 2026
Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)Photo by SimWorld (@simworld_ai)
SimWorld retweeted
Qwen4 Max on BuildingBench ๐Ÿ›๏ธ Photos in, a usable 3D model out. #Qwen #AI #3D #Architecture
2
3
15
995
SimWorld retweeted
Enactra Agentic Simulation ๐Ÿค–๐ŸŽฌ Robots you can control, in worlds you can direct. Every frame comes annotated. ๐Ÿ”Š Sound on #AgenticSimulation #Robotics #EmbodiedAI
22
31
7
135
21,855
SimWorld retweeted
World simulation with both physical & visual fidelity will be playing a key role in physical AI
Enactra Agentic Simulation ๐Ÿค–๐ŸŽฌ Robots you can control, in worlds you can direct. Every frame comes annotated. ๐Ÿ”Š Sound on #AgenticSimulation #Robotics #EmbodiedAI
1
3
14
939
SimWorld retweeted
I've watched this video more than 50 times, always with the sound on. The cinematic quality still impresses me. I highly recommend experiencing it with audio. ๐Ÿ”Š
Enactra Agentic Simulation ๐Ÿค–๐ŸŽฌ Robots you can control, in worlds you can direct. Every frame comes annotated. ๐Ÿ”Š Sound on #AgenticSimulation #Robotics #EmbodiedAI
2
12
772
This is a massive game-changer for Embodied AI! Having a highly controllable simulation where "every frame comes annotated" out-of-the-box is going to save researchers countless hours of data processing. The visual fidelity is absolutely incredible too.! #EmbodiedAI #Robotics
Enactra Agentic Simulation ๐Ÿค–๐ŸŽฌ Robots you can control, in worlds you can direct. Every frame comes annotated. ๐Ÿ”Š Sound on #AgenticSimulation #Robotics #EmbodiedAI
12
18
112
16,615
Can reinforcement learning solve inverse kinematics precisely? Introducing ๐—ฃ๐—ผ๐—น๐—ถ๐—ฐ๐˜†๐—œ๐—ž ๐Ÿฆพ, a goal-conditioned RL policy for high-precision inverse kinematics. On a real arm, it reaches median errors of ๐Ÿฌ.๐Ÿฑ๐Ÿฌ ๐—บ๐—บ ๐—ฎ๐—ป๐—ฑ ๐Ÿฌ.๐Ÿญ๐Ÿฎยฐ and holds the goal pose on ๐Ÿฎ.๐Ÿฑร— ๐—บ๐—ผ๐—ฟ๐—ฒ ๐˜๐—ฎ๐˜€๐—ธ๐˜€ than the best classic IK controller. ๐ŸŒ tinaliu0123.github.io/Policyโ€ฆ ๐Ÿงต๐Ÿ‘‡
15
57
10
507
52,750
SimWorld retweeted
"No OCR, read text natively from images" continues! We showed this with deep research agents: it not only improves performance but also reduces tool calls! Post-training to work with page images gives a big boost - likely because these open-weight models haven't been trained to work with page images yet? Exciting work led by @Hughie_Hu43 !
(1/7) ๐Ÿš€ Introducing DR Pixel, the first open-source deep research agent that searches and reasons directly over page images, no OCR. On 1.06M pages, our trained 27B model matches GPT-5.4. ๐Ÿ“„ arxiv.org/abs/2611.03097 Joint work from @BerkeleyNLP, @rtnldynamics!
7
26
2
247
17,858
SimWorld retweeted
Check out SceneBloom! Diffusion-powered, agentic 3D indoor generation for embodied learning.
Can generated scenes train embodied agents better than real-world data? Introducing ๐—ฆ๐—ฐ๐—ฒ๐—ป๐—ฒ๐—•๐—น๐—ผ๐—ผ๐—บ ๐ŸŒธ, an agentic framework that generates realistic, diverse, simulation-ready 3D scenes for embodied learning. ๐Ÿค– ๐—•๐—ฒ๐˜๐˜๐—ฒ๐—ฟ ๐˜๐—ฟ๐—ฎ๐—ถ๐—ป๐—ถ๐—ป๐—ด ๐—ฑ๐—ฎ๐˜๐—ฎ: 1.7ร— the navigation gain of an equal number of real scans โœ… ๐—ฃ๐—ต๐˜†๐˜€๐—ถ๐—ฐ๐—ฎ๐—น๐—น๐˜† ๐—ฝ๐—น๐—ฎ๐˜‚๐˜€๐—ถ๐—ฏ๐—น๐—ฒ: no detected mesh collisions, and the most physically stable ๐Ÿ  ๐—ฅ๐—ฒ๐—ฎ๐—น๐—ถ๐˜€๐˜๐—ถ๐—ฐ: layouts closest to real rooms ๐Ÿ“„ arxiv.org/abs/2610.32978 ๐Ÿ’ป github.com/Tinaliu0123/sceneโ€ฆ ๐ŸŒ tinaliu0123.github.io/scenebโ€ฆ ๐Ÿงต๐Ÿ‘‡
2
6
37
2,324
SimWorld retweeted
RT-SAFE: Are the Newer Agents Safer in REAL TIME Environment? ๐Ÿ‘€ REASONING LATENCY matters โ€” the world keeps moving during inference. โšก DeepSeek V4 Flash Vision โ†’ V4.1 Flash: 50% lower response time; safety events 27.6 โ†’ 20.4 (โˆ’26%). ๐Ÿงญ Gemini 3.8 Flash โ†’ 4 Argon: 30% lower response time, 18% fewer decisions; safety events 16.0 โ†’ 10.4 (โˆ’35%). ๐Ÿ’ญ Kimi K3 โ†’ K3.1 Pro: 33% fewer decisions; safety events 36.0 โ†’ 28.0 (โˆ’22%). Responses still averaged ~42 s. ๐Ÿค– GPT-5.6 Sol โ†’ 6.1 Sol: 11% fewer decisions, but safety events 13.2 โ†’ 16.0 (+21%). โš ๏ธ SAFETY FAILURES PERSIST. Faster or more efficient doesnโ€™t automatically mean safer. More model examples & failure analyses coming soon! ๐Ÿ’ป Code: github.com/XTRose29/RT-SAFE ๐Ÿ“„ Paper: arxiv.org/abs/2610.09294 ๐ŸŒ Website: xtrose29.github.io/RT-SAFE/
2
1
14
1,406
Can generated scenes train embodied agents better than real-world data? Introducing ๐—ฆ๐—ฐ๐—ฒ๐—ป๐—ฒ๐—•๐—น๐—ผ๐—ผ๐—บ ๐ŸŒธ, an agentic framework that generates realistic, diverse, simulation-ready 3D scenes for embodied learning. ๐Ÿค– ๐—•๐—ฒ๐˜๐˜๐—ฒ๐—ฟ ๐˜๐—ฟ๐—ฎ๐—ถ๐—ป๐—ถ๐—ป๐—ด ๐—ฑ๐—ฎ๐˜๐—ฎ: 1.7ร— the navigation gain of an equal number of real scans โœ… ๐—ฃ๐—ต๐˜†๐˜€๐—ถ๐—ฐ๐—ฎ๐—น๐—น๐˜† ๐—ฝ๐—น๐—ฎ๐˜‚๐˜€๐—ถ๐—ฏ๐—น๐—ฒ: no detected mesh collisions, and the most physically stable ๐Ÿ  ๐—ฅ๐—ฒ๐—ฎ๐—น๐—ถ๐˜€๐˜๐—ถ๐—ฐ: layouts closest to real rooms ๐Ÿ“„ arxiv.org/abs/2610.32978 ๐Ÿ’ป github.com/Tinaliu0123/sceneโ€ฆ ๐ŸŒ tinaliu0123.github.io/scenebโ€ฆ ๐Ÿงต๐Ÿ‘‡
2
9
1
60
5,757
SimWorld retweeted
๐Ÿ”ฅSynthetic 3D worlds can beat real-world scans! Introducing SceneBloom: diffusion-powered, agentic 3D indoor generation for training embodied AI. Realistic. Diverse. Physically valid. Executable. โ†’ Better 3D worlds โ†’ Better embodied agents. 1.7ร— navigation gain vs. equal real scans Zero detected mesh collisions 37.45% โ†’ 4.25% contract violations
8
14
1
101
6,467
SimWorld retweeted
Iโ€™m so glad to see open-source coding agents really catching up in 3D coding!
Wow! Kimi K3.1 Pro leaps into BuildingBenchโ€™s top six! Another open model making serious strides in 3D ๐Ÿ—๏ธ๐Ÿ”ฅ ๐ŸŽˆ54.0 โ†’ 81.6 from Kimi K3โ€”a massive 27.6-point jump! ๐ŸŽˆEssentially matches DeepSeek V4.1 Flash 1017โ€™s 81.9, joining the cluster around Sonnet 5.5 and Fable 5.1. ๐ŸŽˆEstimated median cost: $12.58/task. Cheaper than Sonnet and Fable, though DeepSeek remains more cost-efficient at $3.01. The reconstruction shown here scores 90.8: Melbourneโ€™s Australian Centre for Contemporary Art, rebuilt from photographs. Exciting to see more open models reaching this level of spatial reasonin!
1
3
30
1,585
SimWorld retweeted
Wow! Kimi K3.1 Pro leaps into BuildingBenchโ€™s top six! Another open model making serious strides in 3D ๐Ÿ—๏ธ๐Ÿ”ฅ ๐ŸŽˆ54.0 โ†’ 81.6 from Kimi K3โ€”a massive 27.6-point jump! ๐ŸŽˆEssentially matches DeepSeek V4.1 Flash 1017โ€™s 81.9, joining the cluster around Sonnet 5.5 and Fable 5.1. ๐ŸŽˆEstimated median cost: $12.58/task. Cheaper than Sonnet and Fable, though DeepSeek remains more cost-efficient at $3.01. The reconstruction shown here scores 90.8: Melbourneโ€™s Australian Centre for Contemporary Art, rebuilt from photographs. Exciting to see more open models reaching this level of spatial reasonin!
10
7
2
44
4,827
Thanks AK @_akhaliq for featuring our work! Excited to share OpenVLHarness, a multimodal agent that inspect, compute, and reason over evidence. The technical core: โ€ข A capability layer gives consistent interfaces for various tools such as perception and search, and renders their outputs into observations the model can inspect. โ€ข Persistent multimodal memory retains images, masks, numerical arrays, and web references under reusable handles. โ€ข A coding agent computes over those artifacts to combine evidence across tools. Across 22 benchmarks and 8 backbones, we see average gains of 9.7โ€“12.6 points over matched no-tool baselines. Evidence acquisition and additional reasoning are complementary! A joy collaborating with this UCLA ร— Apple team. Huge shoutout to @yu_bryan_zhou, @cfyang58, @RuiSun94013021, @DiWu0162 , Sibo Peng, Mingyang Zhang, Zhen Yang, Ruiyi Zhang, @zhegan4 and @kaiwei_chang! ๐ŸŽ‰
OpenVLHarness Real World Visual Agentic Tool-Use with Multimodal Memory paper: huggingface.co/papers/2610.2โ€ฆ
4
8
38
3,028
SimWorld retweeted
MiniMax M3.1 Flash reshapes BuildingBenchโ€™s costโ€“quality frontier! Another open model making moves in 3D ๐Ÿ—๏ธ๐Ÿ”ฅ ๐ŸŽˆ 73.3 overall, ranking #12. ๐ŸŽˆ Above SWE-2โ€™s 67.3, taking its place on the Pareto frontier. ๐ŸŽˆ Close to the previous DeepSeek V4.1 Flashโ€™s 72.9โ€”and currently free through OpenRouter! Its best reconstruction hits 88.8 on Melbourneโ€™s Australian Centre for Contemporary Art. How does it look alongside DeepSeek V4.1 Flash 1017 and #1 Opus 5.5? Comparison video below ๐Ÿ‘‡
4
8
1
27
2,327
SimWorld retweeted
See DeepSeekโ€™s 3D leap with your own eyes ๐Ÿ‘€๐Ÿ—๏ธ DeepSeek V4.1 Flash 1017 ๐Ÿ†š DS V4.1 Flash ๐Ÿ†š GPT-6 Astra: three buildings, side by side. The new open model improves on 11 of 12 buildings in BuildingBench, jumping from 72.9 โ†’ 81.9โ€”closing in on Astraโ€™s 84.3 at less than half the median cost ($3.01 vs $7.05/task). Open models are getting seriously good at building worlds ๐Ÿ”ฅ
Wow๐Ÿ™€๐Ÿ™€ DeepSeek V4.1 Flash 1017 joins BuildingBenchโ€™s top fiveโ€”an exciting leap for open models!! ๐Ÿ”ฅ We tested its 3D building reconstruction from real photos. An open model delivering this level of spatial reasoning at this cost is exciting to see! ๐Ÿ‘€ ๐ŸŽˆ 81.9 overall, up from 72.9 for the previous version. A 9-point jump! ๐ŸŽˆ Slightly higher than Fable 5.1โ€™s 81.4, and essentially level with Sonnet 5.5โ€™s 82.1. ๐ŸŽˆ Estimated median cost: $3.01/task. About 1/11 of Fable 5.1โ€™s cost and 1/15 of Sonnet 5.5โ€™s.
7
11
3
64
7,466
SimWorld retweeted
๐ŸŽ‰Excited to see two new US open models from this week: ๐Ÿ“˜Metaโ€™s Muse Spark Open ๐Ÿ“—NVIDIAโ€™s Nemotron 3.5 Super VL How well can they build in 3D? ๐Ÿ—๏ธ๐Ÿ‘€ We tested both on BuildingBench: reconstructing real buildings from photographs. ๐Ÿ“˜Muse Spark Open (max): 56.9 at an estimated $8.59/taskโ€”comparable to Kimi K3โ€™s 54.0. ๐Ÿ“— Nemotron 3.5 Super VL (max): 37.8 at an estimated $0.76/taskโ€”among the cheapest tested, with performance near GLM-5.3 Flashโ€™s 33.0. DeepSeek V4.1 Flash remains well ahead of both: 72.9 at an estimated $2.33/task. GLM-5.3 FlashX also scores higher, at 68.0. Neither newcomer reaches the costโ€“quality frontier. More open models to choose from, and plenty of room to improve spatial reasoning! Both ran in Claude Code at max effort. Costs are estimated median per task.
5
6
1
17
1,345
SimWorld retweeted
Watch how GPTโ€‘6 Astra + WorldBox tackle MazeBench, from its first action to gem๐Ÿ’Ž #61โ€”32,980 live actions and around 1B tokens. Building bridges with boxes, launching across gaps, solving puzzles that span multiple roomsโ€ฆ all while the map fills in. Remember, test, revise, repeat, and grab that gem. WorldBox offers the supporting layer for Astra.
๐ŸšจNew level unlocked! ๐Ÿšจ We doubled GPTโ€‘6 Astraโ€™s score on MazeBench with WorldBox, our agentic sandbox for spatial reasoning. Astra: 23 gems ๐Ÿ’Ž๐Ÿ’Ž Astra+WorldBox: 61 gems ๐Ÿ’Ž๐Ÿ’Ž๐Ÿ’Ž๐Ÿ’Ž๐Ÿ’Ž MazeBench, by @patience_cave, is a tough test of long-horizon spatial reasoning: 256 interconnected rooms where even Opus 5.5 used 2.2B tokens and collected just 13 of 100 gems. WorldBox provides agents with an agentic sandbox, for managing long-term memory, inferring complex mechanics, and simulating reasoning. Watch the sandbox take shape and the agent tackle increasingly tricky puzzles. ๐Ÿงต
2
4
1
9
733
SimWorld retweeted
Long-horizon spatial reasoning remains a major challenge, even for the strongest frontier AI models. Impressive results from @EnactraAI! ๐Ÿš€ WorldBox boosts GPT-6 Astra from 23 to 61 gems on MazeBench, a 2.7ร— improvement! Exciting to see how memory, world modeling, and simulation can help AI agents tackle much harder problems.
๐ŸšจNew level unlocked! ๐Ÿšจ We doubled GPTโ€‘6 Astraโ€™s score on MazeBench with WorldBox, our agentic sandbox for spatial reasoning. Astra: 23 gems ๐Ÿ’Ž๐Ÿ’Ž Astra+WorldBox: 61 gems ๐Ÿ’Ž๐Ÿ’Ž๐Ÿ’Ž๐Ÿ’Ž๐Ÿ’Ž MazeBench, by @patience_cave, is a tough test of long-horizon spatial reasoning: 256 interconnected rooms where even Opus 5.5 used 2.2B tokens and collected just 13 of 100 gems. WorldBox provides agents with an agentic sandbox, for managing long-term memory, inferring complex mechanics, and simulating reasoning. Watch the sandbox take shape and the agent tackle increasingly tricky puzzles. ๐Ÿงต
1
8
27
1,799
SimWorld retweeted
Hereโ€™s what that progress looks like across the maze. ๐Ÿ—บ๏ธ Rooms visited by the end of each run: ๐Ÿฅ‰ Opus 5.5: 68 ๐Ÿฅˆ GPTโ€‘6 Astra: 113 ๐Ÿฅ‡ GPTโ€‘6 Astra + WorldBox: 230 of 256 Brighter tiles mark more frequent visits. With WorldBox, the agent reaches almost 90% of the rooms and steps onto nearly twice as many distinct tiles as the Astra baseline. WorldBox starts with empty memory and no skills. The agent writes its own notes and Python skills, reconstructs room geometry from camera views, and tests candidate moves in the sandbox. Remembering where youโ€™ve been helps make the next exploration decision. What does that reasoning look like inside a difficult room? ๐Ÿ‘‡
1
3
6
434