@scievision369 how do we know that is the same particle that is leaking on the other side of the wall? It is not like only the incoming object has particles and not the wall - could it be that the wall absorbed one from the incoming object and released one from the otherside?
This little illuminated dragon is very happy about Pretext. He's too busy having fun to care about people's "hot takes" on how "it's not that special."
(This little dragon also only works on desktop right now but maybe I'll do mobile later)
https://t.co/k9FH6p1G0T
In 1998, Japanese Mathematician, Engineer and Machine Learning pioneer Shun-ichi Amari published "Natural Gradient Works Efficiently in Learning (722 Cites in Papers and 9 Cites in Patents)" and made a point that still feels fresh even today
If your loss is L(θ), then the usual update θ̇ = −∇ L(θ) quietly assumes your parameter space is flat.
Amari asked what happens when the model itself has geometry, described by G(θ), the Fisher information.
Then the more natural direction becomes θ̇ = −G(θ)⁻¹∇ L(θ). This is the same loss, but with a different notion of distance.
As the animation shows, ordinary Gradient Descent follows the raw slope and gets dragged around by distortion, while the natural gradient moves in a way that respects the geometry of the statistical model itself.
this guy has 29 models on huggingface at page 2 ranking. no lab behind him. no sponsorship. $2,000 from his own pocket on GPU rentals. he compressed GLM-4.7 to run on a MacBook and quantized Nemotron Super the week it dropped. all public. all free.
nvidia is a trillion dollar company with hundreds of teams but they are not the ones quantizing models middle of the night and pushing them out before sunrise. if nvidia stopped tomorrow their employees stop working. people like @0xSero would not. that is the difference between a paycheck and a mission.
@NVIDIAAI you talk about making AI accessible. the people actually doing it are right here. 29 models deep burning their own compute with no ask except more hardware to keep going. you do not need to build another program. just look at who is already building for you. one GPU to this man would produce more public value than a hundred internal sprints.
i am not asking for charity. i am asking you to invest in someone who already proved it.
#𝗣𝗵𝗮𝗿𝗺𝗮𝗰𝗼𝗸𝗶𝗻𝗲𝘁𝗶𝗰𝘀 (𝗣𝗞) deals with how the body interacts with a 𝘥𝘳𝘶𝘨 from the moment of administration until its complete elimination. When #𝗔𝗜 is modelled for drug administrations, we use these exact Math carefully to make the #Biology act as "guardrails"
Encore.ts looks interesting, especially with leap, but it is nowhere near production. `encore run` errors logs are just one liners, esm module loading fails, request interfaces do not support anything other than simple types ... better to stay away from it for now. #encore#ts
https://t.co/bBnY560qnM Conventional interfaces are designed for transactional interactions: a user asks a question, the system responds. An immersive interface transcends the point-and-click model by offering a spatial, contextual, and navigable environment. #javascript#ux#mvp
https://t.co/rBKPwzFyhK
JSON-RPC 3.0 with support for #streaming requests and responses (making it easy to use #LLM streaming RPC with Websockets for #AI Agent implementations in #typescript )
https://t.co/7uFMdynaEB Highlights the human relationship with tools and the historical patterns of resistance to new technologies, and how #AI relates to these patterns of #Blockchain#Web3.0 #Industry4.0 and #Startups in the #emergingTech
#Blockchain isn’t a luxury; it’s a necessity. It provides the trust, transparency, and security required for #AI to thrive in high-stakes, real-world applications. https://t.co/hjgbjxVYX1 #startup#aiml#javascript#cxo#mvp#openai
AI is a tool, much like a hammer, car or calculator. You wouldn't criticize someone for using a pulley to lift heavy weights.
Yet when someone uses AI to write their mails, reports or drafts, everyone jumps on to shame them.
https://t.co/Q96PHACj2s
RPC methods block in python. Asyncer's asyncify() will use AnyIO underneath to do the smart thing, avoid blocking the main async event loop, and run the sync/blocking function in a worker thread.
https://t.co/MAeoHDf87s #python#mvp#startup#cto#RPCs#ai#Blockchain
Interviews are 2way street, especially when hiring for senior positions. As startup founder, you have to be very mindful of what and how you seek answers from potential candiates.
Long before you get to judge their answers, you will be measured on your questions.
Dont be naive
When OpenAI trained the model on internal Slack messages.
User: "Write a 500 word blog post on prompt engineering"
Assistant: "Sure, I shall work on that in the morning"
User: "Write it now"
Assistant: "ok"
#OpenAIDevDay – fine-tuning on 140k internal #Slack messages #LLM#AI