Matching posts

Matching terms are highlighted

POST

GrokBot praised for simplicity, computer use called the big unlock

…ned his Lia plus OpenClaw setup to be, that its simplicity stands out and computer use is the real …

personal-agent@kavinbm　Filed in No.012

1 day agoCopy original

Similarweb: Muse hits App Store #3, ahead of ChatGPT

Similarweb's report ranks Muse #3 on the App Store,…

muse@Similarweb　Filed in No.012

1 day agoCopy original

NVIDIA publishes 2,000+ SimReady 3D assets on Hugging Face

NVIDIA Robotics released over 2,000 simulation-ready 3D assets from seven robot OEMs on…

3d@lukas_m_ziegler　Filed in No.004

1 day agoCopy original

POST

SenseTime evaluates GPT-6 Astra across 12 sim benchmarks and 8 real robots

…SenseTime_AI just studied GPT-6-Astra across 12 simulation benchmarks and 8 real-robot platforms. T…

gpt-6@DataChaz　Filed in No.001

1 day agoCopy original

AV3D-Harness open-sourced for 3D audio-visual real2sim, inspired by Astra demos

… first agentic harness for 3D audio-visual real2sim. No manual acoustic tuning. AV3D-Harness autono…

gpt-6@lasiafly　Filed in No.001

1 day agoCopy original

FAWNBENCH: Haiku 5.5 vs M3.1, V4.1 and Terra 6.1 at similar prices

…with M3.1, V4.1 and Terra 6.1, picked for their similar AA pricing, in a video rundown of outputs.

claude@fawnunwrap　Filed in No.010

19 hours agoCopy original

Gemini 4 Argon builds an interactive 3D MONIAC simulation with Three.js

…n and used it to generate an interactive 3D web simulation of the 1949 MONIAC water-powered economi…

gemini-4@IamEmily2050　Filed in No.008

1 day agoCopy original

Gemini 4 Argon pairs with Genie 4 to simulate Apollo playing checkers

… Genie 4 as it generates the scene live—a robot simulation experiment, not a physical robot demo.

gemini-4@humanoidsdaily　Filed in No.008

21 hours agoCopy original

Google pairs Genie 4 with Gemini 4 Argon for robotics simulation research

…Gemini 4 Argon in DeepMind robotics research to simulate scenarios in imagined worlds.

gemini-4@Google　Filed in No.008

22 hours agoCopy original

POST

DeepSeek V4.1 Flash on Terminal-Bench 2.0: similar pass rates, 1.6x token gap across harnesses

On self-hosted vLLM, Hermes and Claude Code both hit 75% and DSH 70%; DSH uses 1.23x Hermes prompt tokens while Claude Code uses 0.74x but caches only 9–11% of prefixes.

deepseek-v4.1@nextaurame　Filed in No.020

1 day agoCopy original

Showing10of60posts

10 per batch

Keep scrolling to reveal posts 11–20

Showing the first 60 matches. Narrow the query with a model name or tag to find more specific results.