TheCodingSoup's profile picture. Developer Advocate | AI Solutions Engineer |

Lorenzo Roller

@TheCodingSoup

Developer Advocate | AI Solutions Engineer |

Lorenzo Roller đã đăng lại

Studying math notes used to feel like: \int_a^b f(t)dt.... Now it's: ∫𝑓(𝑡)𝑑𝑡 LaTeX has officially arrived in @NotebookLM! 🎈 All of your equations are now perfectly rendered in Chat, Flashcards, and Quizzes. So that upcoming algebra final should be easy as π.


Lorenzo Roller đã đăng lại

What if scaling the context windows of frontier LLMs is much easier than it sounds? We’re excited to share our work on Recursive Language Models (RLMs). A new inference strategy where LLMs can decompose and recursively interact with input prompts of seemingly unbounded length,…

a1zhang's tweet image. What if scaling the context windows of frontier LLMs is much easier than it sounds?

We’re excited to share our work on Recursive Language Models (RLMs). A new inference strategy where LLMs can decompose and recursively interact with input prompts of seemingly unbounded length,…

Lorenzo Roller đã đăng lại

We are excited to share CleanMARL, a project that provides clean, single-file implementations of Deep Multi-Agent Reinforcement Learning (MARL) algorithms in PyTorch, following the same philosophy as CleanRL. #RL


Lorenzo Roller đã đăng lại

Your RL run just spiked at step 89! But, do you know why? We’re fixing that. Today we’re launching W&B Weave Traces to give you a step by step look into your agent’s decisions. This is the first drop from our fresh new integration with @OpenPipeAI. More RL magic is incoming.


Lorenzo Roller đã đăng lại

Announcing "look at your damn data" for Reinforcement Learning! You now get the full power of @weave_wb traces in the @weights_biases workspace. For any wandb run that calls an LLM, just "pip install weave" and "import weave" in your code. We also have out-of-the-box…


Lorenzo Roller đã đăng lại

Seeking students & open-source contributors to join some ML-for-science projects at @huggingface. If you're curious about ML × biology or ML × materials science, this could be a great way to learn + contribute. I can offer Pro subs, support, and (for longer-term work)…


Lorenzo Roller đã đăng lại

Huge computer science result: A Tsinghua professor JUST discovered the fastest shortest path algorithm for graphs in 40yrs. This improves on Turing award winner Tarjan’s O(m + nlogn) with Dijkstra’s, something every Computer Science student learns in college.

deedydas's tweet image. Huge computer science result:

A Tsinghua professor JUST discovered the fastest shortest path algorithm for graphs in 40yrs.

This improves on Turing award winner Tarjan’s O(m + nlogn) with Dijkstra’s, something every Computer Science student learns in college.

Lorenzo Roller đã đăng lại

Announcing MCP•RL: teach your model how to use any MCP server automatically using reinforcement learning! Just connect any MCP server, and your model will start playing with it and (using RL) "learn from experience" how to use its tools most effectively!

corbtt's tweet image. Announcing MCP•RL: teach your model how to use any MCP server automatically using reinforcement learning!

Just connect any MCP server, and your model will start playing with it and (using RL) "learn from experience" how to use its tools most effectively!

Lorenzo Roller đã đăng lại

thank you to our partners at microsoft, nvidia, oracle, google, and coreweave for making this possible! lots and lots of GPUs working overtime.


Lorenzo Roller đã đăng lại

NEW: Kimi K2 is now live on W&B Inference by @CoreWeave! It's the first truly open challenger, ready for production with no waitlists, no restrictions, and SOTA performance. Here's why we think it's larger than the DeepSeek moment and how to claim $50 in inference credits. 🧵


Lorenzo Roller đã đăng lại

Introducing our latest technical report: Context Rot - How Increasing Input Tokens Impacts LLM Performance Our results reveal that models do not use their context uniformly. full report in replies


Lorenzo Roller đã đăng lại

The single most undervalued fact of linear algebra: matrices are graphs, and graphs are matrices. Encoding matrices as graphs is a cheat code, making complex behavior simple to study. Let me show you how!

TivadarDanka's tweet image. The single most undervalued fact of linear algebra: matrices are graphs, and graphs are matrices.

Encoding matrices as graphs is a cheat code, making complex behavior simple to study.

Let me show you how!

Lorenzo Roller đã đăng lại

This paper proposes Agentic Neural Networks (ANN). ANN treats multi-agent collaboration as a layered neural network. It automatically builds and improves agent teams and their interactions using data-driven methods. Methods 🔧: → Tasks decompose dynamically into layers. →…

rohanpaul_ai's tweet image. This paper proposes Agentic Neural Networks (ANN).

ANN treats multi-agent collaboration as a layered neural network.

It automatically builds and improves agent teams and their interactions using data-driven methods.

Methods 🔧:

→ Tasks decompose dynamically into layers.

→…

Lorenzo Roller đã đăng lại

Introducing your arXiv Research Agent A personal research assistant with access to arXiv + bioRxiv + medRxiv + Semantic Scholar. Upload drafts, conduct literature reviews, get insights across millions of papers MCP support coming soon 🚀


Lorenzo Roller đã đăng lại

much of the AI internet is down now firebase is down, cursor is down, lovable is down, supabase is down, google ai is down, cursor is down, aws is down... almost everything is down. finally time to catch up on the 87 tools, 14 models, and 12 AI startup ideas we want to build.


Lorenzo Roller đã đăng lại

New Claude Code SDKs (TypeScript + Python) are here, letting you embed Claude Code capabilities directly into your tools.

TypeScript and Python SDKs for Claude Code are here. Use the same agent and tools as Claude Code, and build them into your own agents and workflows.



Lorenzo Roller đã đăng lại

New in Claude Code: Plan mode. Review implementation plans before making changes. Perfect for complex changes where you want to nail the approach before diving in.


Lorenzo Roller đã đăng lại

npm create mastra

Mastra (@mastra_ai) is the open-source JavaScript framework for building agents. Companies use Mastra to automate support, build CAD diagrams, scrape the web, do medical transcription, and more. ycombinator.com/launches/Muk-m… Congrats on the launch @calcsam, @smthomas3 and @abhiaiyer!



Lorenzo Roller đã đăng lại

You weren't dreaming— the @NotebookLM mobile app started rolling out this morning! We were eager to get the app into your hands, so this initial version has an MVP feature set with more functionality coming soon! Here are a few of the features we're most excited about: 🧵🧵🧵


Lorenzo Roller đã đăng lại

GitHub Coding Agent announced, built on GitHub actions and deeply integrated with the GitHub UI. Here's the process: - Create a detailed issue - Assign the issue to Copilot - Get a PR with the code changes - Review the code in GitHub UI, request changes etc - Profit

mattpocockuk's tweet image. GitHub Coding Agent announced, built on GitHub actions and deeply integrated with the GitHub UI.

Here's the process:

- Create a detailed issue
- Assign the issue to Copilot
- Get a PR with the code changes
- Review the code in GitHub UI, request changes etc
- Profit

United States Xu hướng

Loading...

Something went wrong.


Something went wrong.