🌱 Paper Finder is a step toward a fully agentic scientific assistant. Join us on the journey. 💬 Discuss on Discord: discord.gg/SyY85E97M5 📚 Learn more about Paper Finder: allenai.org/blog/paper-fin… 💻 Code: github.com/allenai/asta-p…

🍌"Google Goes Bananas – A Deep Dive into Gemini 2.5 Flash Image!" I wrote a short piece, where I review Google DeepMind newest #Gemini 2.5 Flash Image model and its core capabilities including where it shines and its current limitations. 👉Read it now: aashi-dutt3.medium.com/google-goes-ba…

How to turn any PDF into an interactive research assistant with DeepSeek V3.1 + Streamlit ! Extract & parse papers of any length with quick Q&A or deep reasoning with “think”/“non-think” modes 👉 datacamp.com/tutorial/deeps… Powered by OpenRouter

🚨 New model alert! K2-Think (not Kimi) from UAE's MBZUAI is the best open model for Math and Reasoning! This 32B Beast is crushing much larger models like Deepseek R1 and Qwen-235B.

Video and blog post of Songlin Yang explaining DeltaNet used in the newest Qwen-Next model * sustcsonglin.github.io/blog/2024/delt… * youtu.be/d0HJvGSWw8A

🚀Sharing my 2 latest works with Turing College 1. Context Engineering Guide in 2025: turingcollege.com/blog/context-e… 2. Evaluating AI Agents in 2025: A Practical Guide: turingcollege.com/blog/evaluatin…

🆕 Blog: #Qwen3-Next Side-by-side Streamlit demo comparing Qwen3-Next-80B-A3B vs Qwen3-30B-A3B on QA, reasoning, and ultra-long context tasks. 🚀 Fast, efficient, 256K+ context, hybrid MoE 🔗datacamp.com/tutorial/qwen3…