musnit

It's been a while - we've got something new: Spinamp Channels: A dynamic, blazing fast way to discover new artists and collect their music fast, all on mobile We've worked on this a while and we're sure you'll have a ton of fun with it. This launch is filled to the brim:

Can GPT-4 teach a robot hand to do pen spinning tricks better than you do? I'm excited to announce Eureka, an open-ended agent that designs reward functions for robot dexterity at super-human level. It’s like Voyager in the space of a physics simulator API! Eureka bridges the gap between high-level reasoning (coding) and low-level motor control. It is a “hybrid-gradient architecture”: a black box, inference-only LLM instructs a white box, learnable neural network. The outer loop runs GPT-4 to refine the reward function (gradient-free), while the inner loop runs reinforcement learning to train a robot controller (gradient-based). We are able to scale up Eureka thanks to IsaacGym, a GPU-accelerated physics simulator that speeds up reality by 1000x. On a benchmark suite of 29 tasks across 10 robots, Eureka rewards outperform expert human-written ones on 83% of the tasks by 52% improvement margin on average. We are surprised that Eureka is able to learn pen spinning tricks, which are very difficult even for CGI artists to animate frame by frame! Eureka also enables a new form of in-context RLHF, which is able to incorporate a human operator’s feedback in natural language to steer and align the reward functions. It can serve as a powerful co-pilot for robot engineers to design sophisticated motor behaviors. As usual, we open-source everything! Welcome you all to check out our video gallery and try the codebase today: eureka-research.github.io/ Paper: arxiv.org/abs/2310.12931 Code: github.com/eureka-research/E… Deep dive with me: 🧵

1/ Introducing Spinamp Collect: a way for fans and collectors to support artists through minting and collecting tracks directly on desktop and mobile browsers. mirror.xyz/spinamp.eth/kXjYZ…

How many peeps have played with logit bias / how useful have you found it? I think I'm going to start using it to try to improve the reliability of structured outputs.