fun fact about your brain (coming from a psych student): approaching ur studies and exams with a more positive mindset actually helps ur brain to strengthen its’ neural networks, therefore making it easier for u to remember what you studied. more info down below 👇🏻👇🏻👇🏻

The single most undervalued fact of mathematics: mathematical expressions are graphs, and graphs are matrices. Yes, I know. You already heard this from me, but hear me out. Viewing neural networks as graphs is the idea that led to their success.

Business Applications of Neural Networks: https://t.co/waSZeyeWZS

Nature Neuroscience Prediction of neural activity in connectome-constrained recurrent networks https://t.co/go2NFmUvke

Neural networks were conceptually invented in 1943 by neurophysiologist Warren McCulloch and logician Walter Pitts. They proposed a simple mathematical model of a biological neuron that could take binary inputs and "fire" if the sum exceeded a threshold. They proved that networks

In 2030, neural networks will have approximately the same number of connections as the human brain. This does not necessarily mean that something new will be achieved as a result, but I always keep in mind the dialectical relationship between quantity and quality.

A popular curated list of resources to get into LLMs (>65k GitHub ⭐) The LLM course is divided into 3 parts: 🧩 LLM Fundamentals: Covers basics about Maths for ML, Python, and Neural Networks. 🧑‍🔬 The LLM Scientist: Focuses on LLM architectures and the techniques to build

blog describing the need for width in neural networks for manifold seperation, this has been my most beautiful read in ml so far.

Summary chart of deep learning neural networks

How do humans learn internal object representations? To address this, our new paper in Neural Networks compared human object similarity judgments with various DNNs, using unsupervised alignment to evaluate their fine- and coarse-level correspondences. https://t.co/uHA0qSEppW

ChatGPT is NOT all of AI. Vortex-Net, where neural networks meet whirlpools, is one example of what's possible beyond just predicting the next token. This method opens doors to physics. #AI #MachineLearning

Group theory is the mathematical study of symmetry. In machine learning, it's the foundation for Geometric Deep Learning. Its principles ensure models respect the data's structure. For example, Convolutional Neural Networks (CNNs) are "translation equivariant," meaning they

We are physicists, neuroscientists, mathematicians and computer scientists who study intelligence in biological and artificial neural networks and use our findings to build better AI models. 👉You can join us via the ECE, CS, Math or DYNS programs!

Who Said Neural Networks Aren’t Linear?? In this paper, authors are able to collapse the training of diffusion models down to only 1 step by introducing Linearizer, which sandwiches a linear matrix A between two invertible neural networks!

we’ve already hit the silicon wall. electrons can only move so fast, capping how quickly GPUs can train AI. the next leap? LIGHT. optical neural networks use light waves instead of electricity, when light interferes, it computes. matrix multiplication at the speed of photons.

@SentientAGI RESEARCHER SPOTLIGHT: Edoardo Contente One of the earliest researchers at #Sentient, Edoardo’s journey began at Princeton, where he studied Physics and explored the deep theory behind how neural networks learn and adapt. #SentientAI #AIResearch

1/ Google strikes again: They have introduced DeepSomatic, an AI model that detects genetic mutations in tumor cells by analyzing sequencing data with convolutional neural networks. It outperforms existing tools in accuracy and adaptability, helping researchers identify

Renormalization group for deep neural networks: Universality of learning and scaling laws.

Big updates ahead from $BIO. We have new projects on the BIO launchpad, DAO, IPT and agent. First one is Neuron-Gale. This project is all about harnessing neural science and AI to map brain networks in ways that communities can contribute and benefit from. If you’ve ever