For a limited time, GPT-5 is default in anycoder using @poe_platform https://t.co/ucluakTjAL

InfiGUI-G1 Advancing GUI Grounding with Adaptive Exploration Policy Optimization https://t.co/YV2TQlC0Ho

Vibe coding with @xai grok 4 and @Alibaba_Qwen Image on my phone https://t.co/l7mb3klL2N

Voost A Unified and Scalable Diffusion Transformer for Bidirectional Virtual Try-On and Try-Off https://t.co/8LTe8FPp16

GLM-4.5 Agentic, Reasoning, and Coding (ARC) Foundation Models https://t.co/Wpy7bXJjva

https://t.co/kpPWZwwrnX

Big news for ai in New York https://t.co/50MaR2dHKn

I want to make a rant as a paper too https://t.co/zncGaEARW3

“Write a single paragraph about someone who doles out their remaining words like wartime rations, having been told they only have ten thousand left in their lifetime. They’re at 47 words remaining, holding their newborn.” https://t.co/Q7AMO0436Z

+2 more

“Claude Opus 4.1, De-Carcinize the Great Gatsby” (That was the only prompt) Pretty clever, actually. https://t.co/KbBZE6ur7a

+2 more

Apple's iPad site is running Next.js while all other pages aren't even running React. Is this new? https://t.co/n9VkHAys32

https://t.co/hye9OhZI9c

Agentic Document Extraction now supports field extraction! Many doc extraction use cases extract specific fields from forms and other structured documents. You can now input a picture or PDF of an invoice, request the vendor name, item list, and prices, and get back the extracted fields. Or input a medical form and specify a schema to extract patient name, patient ID, insurance number, etc. One cool feature: If you don't feel like writing a schema (json specification of what fields to extract) yourself, upload one sample document and write a natural language prompt saying what you want, and we automatically generate a schema for you. See the video for details!

I'm thrilled to announce the definitive course on Claude Code, created with @AnthropicAI and taught by Elie Schoppik @eschoppik. If you want to use highly agentic coding - where AI works autonomously for many minutes or longer, not just completing code snippets - this is it. Claude Code has been a game-changer for many developers (including me!), but there's real depth to using it well. This comprehensive course covers everything from fundamentals to advanced patterns. After this short course, you'll be able to: - Orchestrate multiple Claude subagents to work on different parts of your codebase simultaneously - Tag Claude in GitHub issues and have it autonomously create, review, and merge pull requests - Transform messy Jupyter notebooks into clean, production-ready dashboards - Use MCP tools like Playwright so Claude can see what's wrong with your UI and fix it autonomously Whether you're new to Claude Code or already using it, you'll discover powerful capabilities that can fundamentally change how you build software. I'm very excited about what agentic coding lets everyone now do. Please take this course! https://t.co/HGM8ArDalK

OpenAI GPT-5 System Card released "GPT-5 is a unified system with a smart and fast model that answers most questions, a deeper reasoning model for harder problems, and a real-time router that quickly decides which model to use based on conversation type, complexity, tool needs, and explicit intent"

PSA: If you say "think deeply" then you get the thinking model in ChatGPT for free. If you click "ChatGPT Thinking", it costs $20/month min to access, and you get limited usages. https://t.co/v18lPLHbtx

Looks like Deep Research is still using an older model? GPT 5 knows about FastHTML (which is a year old), but if I select "Deep research" then it doesn't. https://t.co/qd3My9NSOU

LexiconTrail shows how to build 10x faster agentic AI systems using @nvidia Small Language Models with our advanced indexing capabilities. This open-source project demonstrates a production-ready architecture that combines specialized SLMs with LlamaIndex's semantic search, knowledge graphs, and multi-modal indexing: 🚀 Multi-agent orchestration with intelligent routing between specialized SLMs for different tasks ⚡ 90% reduction in computational resources while achieving superior accuracy across document QA and reasoning tasks 🏗️ Complete integration with our VectorStoreIndex, KnowledgeGraphIndex, and custom QueryEngine implementations 📊 Real benchmarks showing 240ms response times vs 2400ms for traditional LLM approaches The system leverages our TreeIndex for structured navigation, DocumentSummaryIndex for hierarchical processing, and ResponseSynthesizer with custom prompts - proving that smaller, specialized models can outperform larger ones when properly orchestrated. Built by The AI Cowboys! Explore the repository: https://t.co/hxYo6UioGu

Using @claudeai you now have search results as content blocks, bringing citations to agent applications - no more document workarounds needed! 𝙎𝙚𝙖𝙧𝙘𝙝 𝙧𝙚𝙨𝙪𝙡𝙩𝙨 𝙖𝙨 𝙘𝙤𝙣𝙩𝙚𝙣𝙩 𝙗𝙡𝙤𝙘𝙠𝙨 by @AnthropicAI enables proper source attribution for results from tool calls, matching the citation quality you get from web search functionality: 🔗 Natural citations with source and title attribution linked back to specific tool calls ⚡ Available on Claude Opus 4.1, Claude Sonnet 4, and other latest models via Anthropic and Google Vertex AI 🛠️ We've already integrated this into LlamaIndex with full support for citable tool results and agent workflows Read the official docs: https://t.co/hZGphHQ6O6 Get started with LlamaIndex integration: https://t.co/084wRtNOeF

How @11xAIbuild built Alice, the AI SDR 🚀 Onboarding SDRs can take a long time, 11x shrunk this time to days by solving a critical challenge: getting AI to understand complex company materials like humans do. 𝗧𝗵𝗲 𝗯𝗿𝗲𝗮𝗸𝘁𝗵𝗿𝗼𝘂𝗴𝗵: Multi-modal document ingestion and parsing using LlamaParse ✅ PDFs, PowerPoints, and all sorts of documents - all parsed automatically and made legible to LLMs ✅ Fine-grained parsing control for tables, images, and unstructured data LlamaParse's ability to handle diverse file types with accuracy, plus developer-first tools that let 11x focus on building their agent, not parsing infrastructure. Want the full technical breakdown? 📖 Read the case study: https://t.co/2ulmX71VNc 🎥 Watch the technical deep dive from the team at the latest @aiDotEngineer : https://t.co/xyHIg4nXin

I'm going all in https://t.co/4TEG1SS72z

Maximizing Confidence Alone Improves Reasoning "In this paper, we propose RENT: Reinforcement Learning via Entropy Minimization – a fully unsupervised RL method that requires no external reward or ground-truth answers, and instead uses the model’s entropy of its underlying… https://t.co/N1uHolRGYJ

New Lens on RAG Systems RAG systems are more brittle than you think, even when provided sufficient context. Great work from Google and collaborators. Good tips for devs included. Here are my notes: https://t.co/ip4IoajAsp

Whoever did this paper really should redo it, this is clearly wrong baselines https://t.co/W8EsTpZBoe

Something interesting I just noticed in Hermes' cots during RL - It decided to simulate a python interpreter. The models yearn for the tools.. https://t.co/5GqHPrMXaC

A love letter to @claude_code: Over the past two days, I built a stock analysis web app for loved ones that includes auth, charting tools, stock data/llm apis, db persistance, and more. (see 1 min demo) The velocity was only possible with claude code—please try it if you haven't https://t.co/nNqoCMy0Sv

lol got @AmpCode to create its own branch and commits love it https://t.co/mipxCujp4x

Enhanced structured output support for @openai, now in LlamaIndex! @openai recently expanded their structured output support to include new data types like Arrays, Enums and more, plus string constraint fields like dates and times, emails, and IP addresses. We support all of… https://t.co/SlkVrMmzRA

On two of the most common tests of creativity (the DAT and the AUT), recent models scored well above the average human in creativity, but not as high as the most creative humans There was lots of variability, but better prompts seem to improve performance https://t.co/FQw1ZjI4Up https://t.co/dBtktkkuTb

Crime has declined significantly since its peak in the early ’90s. Yet today, 63% of Americans say crime in the U.S. is 'extremely' or 'very serious,' a 7-point increase from 1996, when actual crime rates were nearly twice as high. https://t.co/hu91uSsBss

You don’t need frontier lab resources for frontier lab automated LLM evaluation. To prove this, we’re open-sourcing j1-nano and j1-micro: two absurdly tiny (600M & 1.7B parameters) but mighty reward models competitive with orders-of-magnitude larger peers. j1-nano and j1-micro… https://t.co/HvMLJMKpvF

"can you do the same thing for football?" - I got this question 100x today yep, I actually did this last year - player detection and tracking - team clustering - perspective transformation end-to-end YT tutorial: https://t.co/d09LtEKz2C https://t.co/Gdg1XS7PFN