Amir Zeinali
@amir_zeinali
Followers
118
Following
5K
Media
55
Statuses
445
Machine Learning Consultant | AI/ML, GenAI, MLOps | Problem Solver
Canada
Joined February 2013
Manus AI, a Chinese AI marvel stunned with a professional Tesla stock analysis in just one hour: .proof it’s the DeepSeek-like game-changer in AI, sparking global buzz and rivalry in 2025.
Manus, the new AI product that everyone's talking about, is worth the hype. This is the AI agent we were promised. Deep Research+Operator+Computer Use+Lovable+memory. Asked it to "Do a professional analysis of Tesla stock " and it did ~2wks of professional-level work in ~1hr!
0
0
2
Sebastian Raschka put together a solid breakdown of 12 new research papers on LLM reasoning. He goes through different ways models can get better at reasoning without needing massive training runs. If you're curious about where things are headed, this is worth checking out @rasbt.
I just shared a new article, "The State of Reasoning Models", where I am exploring 12 new research articles on improving the reasoning capabilities of LLMs (all published after the release of DeepSeek R1): 1. S1: Simple test-time scaling.2. Test-Time.
0
0
0
My favorite one was watching the 'Manus' operating video. The AI agent, which launched in China, was automating about 50 tasks and scenarios simultaneously.
Everyone is talking about MCP, but this was a massive week in AI Agents. I summarized everything announced by OpenAI, LangChain, AutoGen, Hugging Face, LlamaIndex, Reworkd, Composio, MetaGPT, & more. Here's everything you need to know & how to make sense of it: .(save for later)
0
1
0
RT @AtomSilverman: Everyone is talking about MCP. I summarized all the MCP announcements/launches from Composio, Firecrawl, Cursor, Jira,….
0
553
0
RT @rasbt: Setting up your code environment is often the 1st step when building LLMs for education, research, or production. Made a new vi….
0
131
0
Claude 3.7 Sonnet: Anthropic's most intelligent model yet, with hybrid reasoning for instant or step-by-step thinking, plus Claude Code for advanced coding—redefining AI assistance for developers!.@AnthropicAI.
Introducing Claude 3.7 Sonnet: our most intelligent model to date. It's a hybrid reasoning model, producing near-instant responses or extended, step-by-step thinking. One model, two ways to think. We’re also releasing an agentic coding tool: Claude Code.
0
0
0
How Transformer² Adapts LLMs in Real-Time: A Visual Breakdown. Check out this diagram from the new Transformer-Squared (Transformer²) paper. It illustrates how this self-adaptive LLM framework dynamically adjusts to new tasks without full retraining. Here’s what’s happening:
0
0
0
RT @Hesamation: one of the coolest LLM visualizations:.> interactive and 3D .> see all matrix operations.> step by step layer processes.> i….
0
89
0
RT @tedx_ai: We now have INSTANT access to all data across X thanks to Grok 3 which has INSANE implications… 💥. Worked with @jelanifuel to….
0
601
0
DeepSeek's dropping 5 open-source repos next week for #OpenSourceWeek! No hype, just real progress on AG. small team, big moves. Expect daily unlocks, no gatekeeping, just garage-energy vibes.
🚀 Day 0: Warming up for #OpenSourceWeek! . We're a tiny team @deepseek_ai exploring AGI. Starting next week, we'll be open-sourcing 5 repos, sharing our small but sincere progress with full transparency. These humble building blocks in our online service have been documented,.
0
0
0
Super stoked about @rosstaylor90.’s General Reasoning launch, an open-source hub for AI reasoning! It’s got 1.5M+ problems, tons of data traces, and invites community input to rival big closed-source models. Big props for pushing personal, transparent AI after the ‘R1 moment.
🎉 Excited to release General Reasoning: a new community resource for building open reasoning models. We’re looking to make personal, open reasoners a reality. Starting with a small step in that direction today!. Read the thread in the quote tweet for details, or my personal.
0
0
1
DeepSeek's NSA is basically a brainy shortcut for AI, sparsing out attention to handle crazy-long contexts fast, like 11.6x faster on big sequences, while still crushing benchmarks. It’s tailor-made for GPUs, and with Grok 3 dropping, it’s sparking some serious AI smackdowns.
🚀 Introducing NSA: A Hardware-Aligned and Natively Trainable Sparse Attention mechanism for ultra-fast long-context training & inference!. Core components of NSA:.• Dynamic hierarchical sparse strategy.• Coarse-grained token compression.• Fine-grained token selection. 💡 With
1
0
1
RT @OpenAI: Today we’re launching SWE-Lancer—a new, more realistic benchmark to evaluate the coding performance of AI models. SWE-Lancer in….
openai.com
Can frontier LLMs earn $1 million from real-world freelance software engineering?
0
870
0
RT @Hesamation: we are analyzing the top papers on @huggingface (~4000 papers mostly related to LLMs) and here is a list of the top 20 auth….
0
37
0
RT @rasbt: It's 2025, and I’ve finally updated my Python setup guide to use uv + venv instead of conda + pip! .Here's my go-to recommendati….
0
281
0
So, Google's Gemini just dropped a cool update where it can now pull up past chats to make its responses more on point and connected. This feature's rolling out for English-speaking Gemini Advanced users on Google One AI Premium, showing how Google's stepping up the AI game to.
Rolling out starting today, you can ask Gemini to consider your past chats to craft its responses. Easily pick up where you left off or have it summarize a previous topic. You can view, edit, or delete any chats you’ve had with Gemini, and see when it’s used. Try it in Gemini
0
0
1
How are GenAI, data regulations, Postgres, DuckDB, and open table formats shaping data engineering in 2025?.Here are the key takeaways from @anna__geller’s piece:. 🔹 GenAI drives efficiency, but revenue impact remains limited.🔹 AI agents are evolving, but reasoning challenges.
0
0
1
RT @emollick: There is a lot of important stuff in this new paper by Anthropic that shows how people are actually using Claude. 1) The task….
0
159
0
The @AnthropicAI Economic Index is a new initiative by Anthropic aimed at analyzing AI's economic impact by examining millions of anonymized conversations with their AI model, Claude.
Today we’re launching the Anthropic Economic Index, a new initiative aimed at understanding AI's impact on the economy over time. The Index’s first paper analyzes millions of anonymized Claude conversations to reveal how AI is being used today in tasks across the economy.
0
0
1