OpenAI releasing new open model in coming months, seeks community feedback

Arc Prize 2025

SpacetimeDB: Multiplayer at the speed of light

ARC-AGI without pretraining

Redis creator disappointed by Sonnet 3.7 for coding [video]

Arc-AGI-2 Preview

OpenAI: Deep Research System Card

Couldn't create Bing search API resource

LSTM: The Comeback Story? [video]

ChatGPT vs. DeepSeek: Crazy Chess [video]

Interview with DeepSeek Founder: We're Done Following. It's Time to Lead

Gemini 2.0 Flash Thinking model

An In-Depth Answer Engine

PaSa: An LLM Agent for Comprehensive Academic Paper Search

Meta: Memory Layers at Scale

Nobel Minds 2024 [video]

AI Startup Perplexity Closes Funding Round at $9B Value

GPU Glossary

Deepmind: Project Mariner

12 Days of OpenAI: Day 3 [video]

This Video is AI Generated! SORA Review [video]

The EU-Inc Policy Proposal

OpenAI and others seek new path to smarter AI as current methods hit limitations

ProemNews – Add scientific context to any news article

The Super Weight in Large Language Models

Golden Age of Building

AMA with OpenAI's Sam Altman, Kevin Weil, Srinivas Narayanan, and Mark Chen

Scaling Laws for Diffusion Transformers

Petavoxel-Scale Human Cortex Reconstructed at Nanoscale Resolution [pdf]

What Kind of Pencil Do You Use? [pdf]

Actix (Rust) vs. Zap (Zig) vs. Stdlib (Zig): Performance Benchmark in Kubernetes [video]

No Priors Ep. 80 – With Andrej Karpathy from OpenAI and Tesla [video]

Fast Multidimensional Matrix Multiplication on CPU from Scratch (2022)

Mistral.rs: Faster than llama.cpp on CUDA

Zuck pod interview: Llama 3.1 405B, Meta's open-source AI beats GPT-4

Projected Language Models: A Large Model Pre-Segmented into Smaller Ones

A Visual Guide to Quantization

Azure Llama 3.1 Benchmarks

Autodiff and Adjoints for Differentiable Physics [video]

llama3.1-405B leaked at 3am the day before on 4chan

Abstraction and Reasoning Corpus via Procedural Example Generation

Promus Ventures: DeepTech Growth Index vs Nasdaq, S&P

An Adaptive Stochastic Gradient Method with Non-Negative Gauss-Newton Stepsizes

APIGen: Automated Generation of Verifiable Function-Calling Datasets

Ladybird Announcement [video]

IISc physicists find a new way to represent 'pi'

Scientists Just Discovered a New Formula for Pi Accidentally [video]

Meta Large Language Model Compiler: Foundation Models of Compiler Optimization

OpenAI: Strategic Content Partnership with Time

Optimizing AI Inference at Character.ai

LiveBench: A Challenging, Contamination-Free LLM Benchmark

Francois Chollet – LLMs won't lead to AGI $1M Prize to find true solution [video]

General Intelligence (2024)

Microsoft will switch off Recall by default after security backlash

A New Chapter for Bartender

Open AI: Scaling and evaluating sparse autoencoders [pdf]

Adaptive Horizon Actor-Critic for Contact-Rich Policy Learning

Meta FAIR: Contextual Position Encoding, Learning to Count What's Important

Google responds to leak: Documentation lacks context

VALL-E 2: Achieves Human Parity in Zero-Shot TTS

Seal Leaderboards–private, expert evaluations of leading frontier models

Reproducing GPT-2 (124M) in llm.c in 90 minutes for $20

14,000 Google Search Ranking Features Leaked

Rabbit Gaslit Me, So I Dug Deeper [video]

$30M AI Is Hiding a Scam [video]

26× Faster Inference with Layer-Condensed KV Cache for Large Language Models

Ex-OpenAI employee reported losing 85% of his family's net worth

Binary Embeddings: All the AI, 3.125% of the Fat

John Schulman (OpenAI Cofounder) – Reasoning, RLHF, & Plan for 2027 AGI [video]

Setapp Mobile brings a promising new iPhone app experience to the EU

Scaling Autoregressive Multimodal Models with Vision, Text, Audio, and Action

A browser that's designed for research

GPT-2 Chatbots Top the Arena with +50 Elo, Strongest Model Ever

OpenAI Spring Update [video]

Meta FAIR: Memory Mosaics

Do Llamas Work in English? On the Latent Language of Multilingual Transformers