This Asia Cup format makes no sense. Why are India & Pak in the same group again? It's a setup to guarantee multiple INDvPAK matches,potentially 3! Same thing happens in World Cups too. Can’t @BCCI request @ICC & ACC to put them in separate groups to minimize the no. of matches?

“Everyone knows” what an autoencoder is… but there's an important complementary picture missing from most introductory material. In short: we emphasize how autoencoders are implemented—but not always what they represent (and some of the implications of that representation).🧵

This is probably one of THE most important paper of the last few months. Small language models are sufficiently powerful, operationally suitable, and economical Agentic tasks. - Phi-2 matches 30 billion models running 15x faster. - Serving a 7 billion parameter small language

High quality math is the secret sauce for reasoning models. The best math data is in old papers. But OCRing that math is full of insane edge cases. Let's talk about how to solve this, and how you can get better math data than many frontier labs 🧵

Paper from OpenAI says hallucinations are less a problem with LLMs themselves & more an issue with training on tests that only reward right answers. That encourages guessing rather than saying “I don’t know” If this is true, there is a straightforward path for more reliable AI.

Massive proposal in this paper. Argues chips that compute with real physics, not strict digital rules, can relieve the AI compute bottleneck. Early prototypes hit 1000x speed and show steep energy cuts. AI demand keeps rising while data center power, training bills, and

A classic paper, collab between @AIatMeta , @GoogleDeepMind , and @NVIDIAAIDev Language models keep personal facts in a measurable amount of “storage”. This study shows how to count that storage—and when models swap memorization for real learning. 📡 The Question Can we

Can AI help understand how the brain learns to see the world? Our latest study, led by @JRaugel from FAIR at @AIatMeta and @ENS_ULM, is now out! 📄 https://t.co/y2Y3GP3bI5 🧵 A thread:

🏓🤖 Our humanoid robot can now rally over 100 consecutive shots against a human in real table tennis — fully autonomous, sub-second reaction, human-like strikes.

🍌 nano banana is here → gemini-2.5-flash-image-preview - SOTA image generation and editing - incredible character consistency - lightning fast available in preview in AI Studio and the Gemini API

India launched a program to build native large language models for its many languages, funding startups and pooling compute under the $1.2 billion IndiaAI Mission. The Ministry of Electronics and Information Technology reserved 19,000 GPUs (including 13,000 Nvidia H100s). So

PyTorch vs TensorFlow: New survey settles the debate. - PyTorch → ~25% faster in training, ~78% faster in eval. Simpler for research. TensorFlow → Stronger in deployment (mobile, browser, server). Read the whole survey here: https://t.co/BHZDfLXdah

so many papers so little time

Introducing DINOv3: a state-of-the-art computer vision model trained with self-supervised learning (SSL) that produces powerful, high-resolution image features. For the first time, a single frozen vision backbone outperforms specialized solutions on multiple long-standing dense

The gpt-oss models from OpenAI are a synthesis of ideas from prior research. Here are 10 interesting papers that were directly used in gpt-oss… (1) Longformer: Introduces sliding window attention, a form of sparse attention that is utilized in alternating layers of both gpt-oss

🏆 We're thrilled to announce that Meta FAIR’s Brain & AI team won 1st place at the prestigious Algonauts 2025 brain modeling competition. Their 1B parameter model, TRIBE (Trimodal Brain Encoder), is the first deep neural network trained to predict brain responses to stimuli

Huge computer science result: A Tsinghua professor JUST discovered the fastest shortest path algorithm for graphs in 40yrs. This improves on Turing award winner Tarjan’s O(m + nlogn) with Dijkstra’s, something every Computer Science student learns in college.

🇮🇳

How will NISAR support agriculture and farming? Hear from the scientists and end users how its data will be used to map crop growth, track plant health, and monitor soil moisture.

🚨 AI models just invented better, novel AI models. Chinese researchers fed all LLM research into a model and it discovered 106 novel AI model architectures that converge to lower loss with better benchmarks. ASI-Arch is one of the coolest AI papers this year. En route AGI.