🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Cerebellum has over 50% of the total number of neurons in the brain: #neuroscience by G. Dunn

Announcing OPT-IML: a new language model from Meta AI with 175B parameters, fine-tuned on 2,000 language tasks — openly available soon under a noncommercial license for research use cases. Research paper & more details on GitHub ⬇️

I'm proud to announce the publication of my first paper as first author! This project, part of my end of master internship, explores unsupervised & hardware-friendly learning in convolutional SNNs.

Are you wondering how large language models like ChatGPT and InstructGPT actually work? One of the secret ingredients is RLHF - Reinforcement Learning from Human Feedback. Let's dive into how RLHF works in 8 tweets!

Looks like the first open source equivalent of ChatGPT has arrived: github.com/lucidrains/PaL… I.e., an implementation of RLHF (Reinforcement Learning with Human Feedback) on top of Google’s 540 billion parameter PaLM architecture

I spent New Years building GPTZero — an app that can quickly and efficiently detect whether an essay is ChatGPT or human written

Data Science Challenge: Obviously this graphic seems wrong but can you explain what common data visualization mistakes are being made here?

I've been studying machine learning for half a decade, and here is the most popular function we use every day: f(x) = max(0, x) Yet many people don't understand a fundamental characteristic of this function. Here is what you need to know:

Our paper making neural networks more modular, sparse and interpretable by penalizing neuron length is now on arXiv. Less is more! Ziming Liu arxiv.org/abs/2305.08746

Has any founder ever had as much impact on their tech ecosystem as Xavier Niel in France? He’s behind: * Station F - Europe’s biggest startup campus * 42 - complete reinvention of CS education * Kima Ventures (most prolific seed VC), NewWave, other funds * now this:

Llama 2 7B chat, running 100% private on Mac, powered by CoreML! ⚡️ We're optimising this setup to get much more faster generation. 🔥

CPU vs FPGA

New Research: Paired competing neurons improving STDP supervised local learning in Spiking Neural Networks frontiersin.org/articles/10.33… #FrontiersIn #Neuroscience

Our latest paper is now published. In this work, we studied classification with STDP-based SNNs. We showed how winner-takes-all competition can naturally promote, in a supervised context, neuron specialization toward target and non-target samples.

Thanks Reviewer 2 😂🥲 #neurIPS24

🎉 See you at #NeurIPS24