Now you can use AI agents to design directly on the Figma canvas, with our new use_figma MCP tool and skills to teach them. Open beta starts today. https://t.co/AQZsFWvvXQ

Want to explore real-world AI-assisted coding with Copilot? GitHub Copilot Dev Days are happening globally. Join local developers to level up your workflow, or apply to host an event for your own community. Find a Dev Day near you. 👇 https://t.co/l6FilTPqyC

New in Claude Code: auto mode. Instead of approving every file write and bash command, or skipping permissions entirely, auto mode lets Claude make permission decisions on your behalf. Safeguards check each action before it runs. https://t.co/kHbTN2jrWw

Perceptio Perception Enhanced Vision Language Models via Spatial Token Generation paper: https://t.co/gRZeg5hPOi https://t.co/UYedoMn2aE

ICE in the airports today https://t.co/H9U38hNLsX

ICE in the airports today https://t.co/H9U38hNLsX

ICML reviews came out, but the website is down lol https://t.co/Igw1vgb8BM

A tragedy is unfolding at AllenAI. Fs in the chat https://t.co/yPwS7uiE1t

A tragedy is unfolding at AllenAI. Fs in the chat https://t.co/yPwS7uiE1t

There will be more of this. And as much as we're joking about it, we're seeing a massive degradation of code quality right now and we're increasingly only catching it way too late. https://t.co/dq2dWQ9bw4

exciting to try out even better streaming mechanics from FastAPI https://t.co/O4dIJVERDc

Omni-WorldBench Towards a Comprehensive Interaction-Centric Evaluation for World Models paper: https://t.co/asR8atev9A https://t.co/rVycyl7PGC

LongCat-Flash-Prover Advancing Native Formal Reasoning via Agentic Tool-Integrated Reinforcement Learning paper: https://t.co/NqSbMUGR1k https://t.co/2sqBEQeHTv

Group3D MLLM-Driven Semantic Grouping for Open-Vocabulary 3D Object Detection paper: https://t.co/8NVynfAm2u https://t.co/RzkYdEKhRk

Generalized Discrete Diffusion from Snapshots paper: https://t.co/7Z9QGHVp25 https://t.co/wgscL41fnI

Thank you to everyone in the community who is testing and using Nemotron models. It's great to see Nemotron-Cascade-2, Nemotron-3-Super and Nemotron-3-Nano trending on HF. The Nemotron team is working hard to incorporate all your feedback into Nemotron 4. And yes, Nemotron 3 Ultra is still on track for release. https://t.co/lkEwmlUng9

Chinese people used to be much shorter than Americans, but thanks to better nutrition, this has changed. Chinese girls are now taller than American girls, and Chinese boys are almost as tall as American boys. But Dutch remain much taller. https://t.co/wrGm7X8D2q https://t.co/tDpddCdACO

A reminder that exponentials eventually turn into s-curves. (Though this one is likely to stop sooner than AI ability gains) https://t.co/QeMenDUvEq

Apologies to @waitbutwhy for modifying his (excellent and fundamentally correct) graph. https://t.co/zxn5gN3iPO

@nagaraj_arvind I'm focused on "how to make AI helpful to humans". We created https://t.co/gXy2ALgLEj to help towards that. It's been fantastic - very happy spending time on that! :)

Designing better AI may mean making it more honest. Researchers are working on systems that openly communicate uncertainty, especially in areas like medical diagnosis where overconfidence can be dangerous. The goal is not just accuracy, but collaboration with humans. In critical decisions, knowing what AI does not know can matter as much as what it gets right. https://t.co/zm7yhCUPBA @MIT

New from the Anthropic Economic Index: how people’s use of Claude changes with experience. Longer-term users are more likely to iterate carefully with Claude, and less likely to hand it full autonomy. They attempt higher-value tasks, and receive more successful responses. https://t.co/yCsrA0bLt9

We find that since November 2025, consumer use has become less concentrated: the top 10 tasks now make up 19% of conversations, down from 24%. We also see a rise in personal queries, and continued convergence in adoption rates in the US. Read more: https://t.co/9UUq3UsnLi

The Multimodal Future: Why Visual Representation Still Matters, by Saining Xie (@sainingxie) https://t.co/m0zisjJqrM

The Multimodal Future: Why Visual Representation Still Matters, by Saining Xie (@sainingxie) https://t.co/m0zisjJqrM

Now available on Hugging Face: hf-mount 🧑‍🚀 The team really cooked, still wrapping my head everything possible but you can do things like: - mount a 5TB dataset as a local folder and query only the parts you need with DuckDB (✅ works) - browse any model repo with ls/cat like it's a USB drive - use a shared read-write bucket as a team drive for ML artifacts - drop the init container that downloads models in your k8s pods - point llama.cpp at a mounted GGUF and run inference (infinite storage??)

More details here: https://t.co/OT0cPOCGO4

Some links 2,112 total packages depend on litellm (https://t.co/T8xWdrBp9g), 1,403 of them directly. Among these are e.g. DSPy, Open Interpreter, PraisonAI, MLflow, langchain-litellm. Primary Sources 1. Original disclosure (GitHub issue): https://t.co/w0j29cVvdq — the detailed technical writeup of the malicious payload, what it steals, how it works 2. BerriAI's official tracking issue: https://t.co/Jc4iIW7Qmr — the team's response and status updates 3. FutureSearch blog (the people who found it): https://t.co/SaX6ClxW8S — how a fork bomb bug in the malware crashed their machine and led to discovery 4. ramimac's full TeamPCP timeline: https://t.co/nfVesVbrnp — the broader campaign: Trivy → Checkmarx → litellm, with exact timestamps and IOCs 5. PyPA advisory: https://t.co/b7mJsYQkB3 — the official PYSEC-2026-2 Analysis & Discussion 6. Hacker News thread (main): https://t.co/t3BPEMVtk5 — includes direct responses from Krrish (litellm maintainer), best real-time discussion 7. Hacker News thread (early): https://t.co/5qxHxf3pvx — the first HN post 8. GitGuardian writeup: https://t.co/uaCmm1TqHO — good analysis of how leaked CI/CD secrets enabled the whole chain 9. Wiz blog on the broader campaign: https://t.co/CsdOgm9Ios — TeamPCP's attack on Checkmarx/KICS, same actor Downstream Impact 10. DSPy issue: https://t.co/sbgN2AWWlP 11. MLflow emergency pin PR: https://t.co/5gQbxU2Iou 12. DSPy emergency pin PR: https://t.co/AYKgJWdkkz Quick Summary for Your Post - 46-minute window of exposure on PyPI (10:39–11:25 UTC, Mar 24) - ~1,400 packages directly depend on litellm, all with open-ended version ranges - Caught because the malware had a bug (fork bomb) that crashed the discoverer's machine - Root cause: Trivy (a vulnerability scanner) was itself compromised, which leaked litellm's PyPI publish token - Attacker (TeamPCP) spammed the GitHub issue with hundreds of bot comments to bury the discussion

+8 more

Image gen in <1s 🤯 You asked for a demo - here's @iamtimdavis showing Flux 2 Dev on the Modular stack. Reach out to us if you're interested! 🚀 But thats not all - video generation is coming ⬇️ https://t.co/O19Dd26OY7

Cats on MAX ... because of course! 😏 https://t.co/FjBZ1xkQnf

Meta, in collaboration with @huggingface and PyTorch Foundation, is launching India’s first OpenEnv AI Hackathon. Participants will build reinforcement learning environments using OpenEnv, contributing directly to how intelligent agents are trained. @Meta What's in it for developers? $30,000 prize pool! Round 1 (Online): March 25 – April 8 Finale (Bangalore): April 25 – 26 (48-hour in-person hackathon at Scaler School of Technology) Learning resources are provided for those without prior RL experience. 🔗 Register by April 3: https://t.co/hlxBjc1dhL #PyTorch #OpenEnv #MetaAI #HuggingFace #OpenSource

New on the Anthropic Engineering Blog: How we use a multi-agent harness to push Claude further in frontend design and long-running autonomous software engineering. Read more: https://t.co/HWvmXk1ykn