𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp

𝚐π”ͺ𝟾𝚑𝚑𝟾

@gm8xx8

145 Following
131956 Followers
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
LONG LIVE OPEN SOURCE
0 reply
9 recasts
29 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
it’s all coming together.
2 replies
0 recast
11 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
The leap from 2% to 25% on FrontierMath is arguably more impressive than the improvements on ARC-AGI. Exciting times.
0 reply
3 recasts
17 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
For those who aren’t familiar, Alec Radford is the first author of key papers on GPT, GPT-2, CLIP, and Whisper, all of which have played a pivotal role in advancing modern AI.
1 reply
0 recast
11 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Alec Radford exits OpenAI.
1 reply
2 recasts
16 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
figure x openai apptronik x deepmind tesla x xai Apptronik Partners with Google DeepMind: https://apptronik.com/news-collection/apptronik-partners-with-google-deepmind-robotics
0 reply
0 recast
7 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
New simulation engine - Cross-platform: natively supports Nvidia/AMD/Apple/Intel GPU/CPU on Windows, MacOS, and Linux. - Open-source: built entirely in Python (ish). - Speed: outperforms GPU-accelerated platforms like Isaac Gym and MJX by 10-80x, reaching ~430,000x real-time simulation speed. - Efficient: trains robotic locomotion policies for real-world application in 26 seconds on a single RTX 4090, processing 380k steps per second, including policy updates. πŸ”—: https://genesis-world.readthedocs.io/en/latest/
0 reply
0 recast
7 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Gemini 2.0 Flash Thinking Experimental Best for: > Multimodal Understanding: Handles tasks involving multiple data types. > Reasoning Capabilities: Excels at solving complex problems. > Coding: Tackles difficult code and math challenges. > Visible Thinking Process: Shows the model’s reasoning step by step. Details: - Pricing: Free for input and output (up to and over 128K tokens). - Rate Limits: 10 requests per minute (RPM). - Latency: Low. - Knowledge Cutoff: August 2024. - Full CoT traces Built on the foundation of the small and powerful Gemini 2.0 Flash model. Available via the Gemini API in Google AI Studio and Vertex AI
0 reply
0 recast
18 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Apple is exploring robot control through Vision Pro. A system using Apple Vision Pro provides real-time virtual robot feedback, enabling high-quality, barehanded data collection for robot imitation learning without physical hardware. A user study showed live feedback improved data quality, with trajectories successfully replayed on physical robots, offering a scalable solution for data collection.
1 reply
2 recasts
16 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Genesis is a high-performance physics platform for robotics, embodied AI, and physical AI applications. It features: - a universal physics engine for simulating diverse materials and phenomena. - a fast, lightweight robotics simulation platform. - a high-speed, photo-realistic rendering system. - a generative data engine that creates data from natural language prompts. - cross platform supporting Nvidia/AMD/Apple/Intel GPU/CPU, Windows, MacOS, & Linuxβ€”natively. The platform combines advanced physics solvers into a unified framework and includes a generative agent for automated data creation. The open-sourced physics engine, built in Python, outpaces GPU-accelerated systems like Isaac Gym and MJX by 10-80x. It achieves ~430,000x real-time simulation speed and trains robotic locomotion policies in 26 seconds on a single RTX 4090. Genesis aims to: - make physics simulation accessible. - unify cutting-edge solvers for high-fidelity virtual environments. - automate data generation to reduce manual effort.
2 replies
5 recasts
19 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
I’ll say it again: Ultimately, every path converges on robotics. Scalable. Adaptable. Collaborative. This is the future.
2 replies
3 recasts
10 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
The stack could be improved, but you get the gist.
0 reply
0 recast
11 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
An AI-powered agent optimized for seamless GPU rental and mgmtβ€”now we cookin https://github.com/HyperbolicLabs/Hyperbolic-AgentKit
1 reply
2 recasts
18 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
xAI announces Aurora nice boost to image gen on 𝕏. https://x.ai/blog/grok-image-generation-release
2 replies
0 recast
4 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
busy week πŸ›¬
0 reply
0 recast
3 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Sora is here (System Card): https://openai.com/index/sora-is-here/
0 reply
0 recast
7 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
day 3
0 reply
0 recast
5 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
0 reply
0 recast
3 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Willow is a SoTA quantum chip that achieves significant advancements in error correction and computation efficiency. It reduces errors as it scales, addressing a long-standing challenge in quantum error correction, and performs tasks in minutes that would take classical supercomputers 10 septillion years. With 105 qubits and leading performance, Willow sets a strong foundation for scalable quantum computing and future real-world applications. LFG. https://blog.google/technology/research/google-willow-quantum-chip/
0 reply
1 recast
18 reactions
𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
looks like we get sora today 😈
0 reply
0 recast
7 reactions