Jesus Cevallos

[Jesus Cevallos @JesusCevallos9]

How America Turned Against the AI Industry? https://t.co/PtJkxDTGRD via @YouTube

[Jesus Cevallos @JesusCevallos9]

Contributing to the research landscape in LM immunisation! https://t.co/xrgxDWGRHl

[Jesus Cevallos @JesusCevallos9]

(...) A new lifestyle is emerging, where we create only what we want and exclude all that we cannot control or know instantly and superficially. This process, by its intrinsic logic, blocks the kind of serene reflection that could lead us to a shared wisdom. (Fratelli tutti, 49)

[Jesus Cevallos @JesusCevallos9]

This is the second time I've tried to "save time" when studying a paper with AI. Unfortunately, it took me longer to realise the AI was hallucinating...
https://t.co/KUUZ2cP4eo
https://t.co/IbYESM3Pc0
...for indepht things, maybe trust your brain a bit more

[Jesus Cevallos @JesusCevallos9]

https://t.co/7SKRQAXZL7

EPIC encounter and DIALOGUE (you don't see that very often) between two beautiful minds! Thanks @TOEwithCurt !!

[Jesus Cevallos @JesusCevallos9]

our agents are value-based an work on a fully-observable MDP!

[Jesus Cevallos @JesusCevallos9]

We did it! 🤓

Four different implementations of fully bootstrapped Deep Active Inference Agents out there!!

OPEN SOURCED!!

Check 'em!!

https://t.co/6TJM3RmB2b

[Jesus Cevallos @JesusCevallos9]

https://t.co/mG4sHxBm3b

[Eric Laukien @222464]

@ID_AA_Carmack @dshandler @DIrmschler While CUDA/OpenCL runs fast, it is difficult to experiment with since writing kernels is a lot more annoying than just running stuff on the CPU with e.g. OpenMP. Also, DL is very slow in general - unconventional architectures can be made much faster. See: https://t.co/yTIBadu6N5

[Jesus Cevallos @JesusCevallos9]

https://t.co/f9eHigOq9s

Tsinghua University strikes again ^_^

[Loic @LoicReco]

OpenAI quietly dropped a 34-page technical manual on building AI agents that 99% of people will never read.

I spent 3 days coding every single pattern they revealed.

Here's the practical guide to autonomous AI agents:🧵

[Caiming Xiong @CaimingXiong]

Multi-Agent Systems (MAS) can outperform single-agent approaches, however designing MAS manually is difficult, especially when LLM preferences differ from human intuition, and manually designed MAS are hard to adapt to new tasks.

🏆 We are introducing MAS-ZERO: Designing Multi-Agent Systems with Zero Supervision 🏆
-🔥 Consistent outperformance of existing automatic MAS methods
-⚡ Beats popular manually designed MAS baselines where other methods fail
-🧠 Sets a new performance-cost frontier across multiple domains

💻 Project Page: https://t.co/Cmr2pQ1Rzg
📄 Paper: https://t.co/7JjF2ZxFiN
🔗 Code: https://t.co/86fMYD3tul
📚 Explore 1,000+ Discovered MAS designs: https://t.co/rYDmar7mIx

[Akshay 🚀 @akshay_pachaar]

Model Context Protocol (MCP), clearly explained:

[Julian Goldie SEO @JulianGoldieSEO]

Gemma 3 + Ollama + N8N + Open Router 🔥

What a model! Great release Google! 👏🏻

- Runs locally via Ollama
- Outperforms Claude Mini & DeepSeek
- Available in 1B, 4B, 12B & 27B sizes
- 128k context window
- Free API access

Save this video it will save you hundreds on AI costs. 🚀

Want the SOP? DM me.

[GitHub Projects Community @GithubProjects]

ANUS is a powerful open-source AI agent framework for automating complex tasks with natural language.

-> No invite code unlike MANUS.

[Sumanth @Sumanth_077]

LLM Engineer Toolkit: A curated list of 120+ LLM libraries for training, fine-tuning, building, evaluating, deploying, RAG, and AI agents!

100% Open Source

[Jesus Cevallos @JesusCevallos9]

Perplexity gone mad XD
https://t.co/8ZQ5j2mgo3

[Sr Carlos ²³²U @CJavierSaldana]

@zhou_xian_

[el.cine @EHuanglu]

RIP Photoshop.

Gemini 2.0 now can edit image by talking to it , it's insane.

Here’s how it works:

[Cerebras @cerebras]

🚨Cerebras Systems + Sandia National Labs have demonstrated training of a 1 trillion parameter model on a single CS-3 system (!)

This is ~1% the footprint & power of an equivalent GPU cluster.