For more information about Aletheia, powered by Gemini #DeepThink, and our works in AI for mathematical and scientific discovery at @GoogleDeepMind and @GoogleResearch, check out our announcement last week nitter.net/lmthang/status/2021631…! Thang Luong (@lmthang) 6 months in, after the IMO-gold achievement, I’m very excited to share another important milestone: AI can help accelerate knowledge discovery in mathematics, physics, and computer science! We’re sharing Two new papers from @GoogleDeepMind and @GoogleResearch that explore how Gemini #DeepThink together with agentic workflows can empower mathematicians and scientists to tackle professional research problems. Some highlights: The first paper built a research agent #Aletheia, powered by an advanced version of Gemini Deep Think, that can autonomously produce publishable math research and crack open Erdős problems. The second paper, built on similar agentic reasoning ideas, helped resolve bottlenecks in 18 research problems, across algorithms, ML and combinatorial optimization, information theory and economics. See the thread for details about the two papers and the joint blog post. — https://nitter.net/lmthang/status/2021631397614731563#m
GENERATIVE AI
-
Perplexity Computer: Unified AI System for End-to-End Project Management
By
–
Introducing Perplexity Computer.
— Perplexity (@perplexity_ai) 25 février 2026
Computer unifies every current AI capability into one system.
It can research, design, code, deploy, and manage any project end-to-end. pic.twitter.com/dZUybl6VkYIntroducing Perplexity Computer. Computer unifies every current AI capability into one system. It can research, design, code, deploy, and manage any project end-to-end.
→ View original post on X — @abhi1thakur, 2026-02-25 16:27 UTC
-
Aletheia solves 6 of 10 FirstProof problems using Gemini DeepThink
By
–
We ran two Aletheia versions (differing only by base model) powered by Gemini #DeepThink. Together, they solved 6/10 problems (2, 5, 7, 8, 9, 10) per majority expert assessments. Full transparency on our FirstProof interpretation and experiments: arxiv.org/abs/2602.21201. Evaluation is extremely hard! Only a handful of experts can even understand these problems. As such, we have conducted our study very carefully! Crucially, our solutions were generated without any human intervention and submitted within the timeframe of the FirstProof challenge. The lead author of FirstProof confirmed that fact in the public Zulip discussion of our solutions icarm.zulipchat.com/#narrow/….
-
Seedream 5 Lite: New Cost-Effective AI Editing Model
By
–
introducing seedream 5 lite.
— KREA AI (@krea_ai) 25 février 2026
this new editing model is similar to nano banana but for a fraction of its cost.
try it now! pic.twitter.com/N8siLNoZWNintroducing seedream 5 lite. this new editing model is similar to nano banana but for a fraction of its cost. try it now!
-
Reflective Test-Time Planning for Embodied LLMs
By
–
Learning from Trials and Errors Reflective Test-Time Planning for Embodied LLMs
-
Query-Focused Memory-Aware Reranker for Long Context
By
–
Query-focused and Memory-aware Reranker for Long Context Processing
-
New Edition: Learn Generative AI, RAG, and AI Agents
By
–
New updated 2nd Edition, by @keithbourne v/ @PacktDataML "Unlocking Data with Generative AI and RAG — Learn AI Agent Fundamentals with RAG-powered Memory, Graph-based RAG, and Intelligent Recall" Get the book here: http://
amzn.to/49zsIkb -
New book: Agentic Architectural Patterns for Multi-Agent AI Systems
By
–
New release from @PacktDataML @PacktPublishing at http://
amzn.to/3MaHy8T "Agentic Architectural Patterns for Building Multi-Agent Systems: Proven design patterns and practices for GenAI, agents, RAG, LLMOps, and enterprise-scale AI systems" 𝕋𝕒𝕓𝕝𝕖 𝕠𝕗 ℂ𝕠𝕟𝕥𝕖𝕟𝕥𝕤: -
News anchors now explain language model distillation better than experts
By
–
Even news anchors now know what language model distillation is. In fact, they explain it better than I do Thanks to Anthropic 😀
→ View original post on X — @skathirmani, 2026-02-25 13:08 UTC
-
Creativity through constraint and surprise
By
–
Let me tell why this works… Creativity = constraint + pressure + surprise. Generic prompts remove all three. This structure injects all three simultaneously. Here.. you are not asking the LLM to be creative, you're building a box so specific that the only way out is through