Building “Auto-Analyst” — A data analytics AI agentic system with DSPy medium.com/firebird-techn…

👋Come to our AI Agents 🤖London meetup We’ll have a deep dive into 2 awesome AI Agent frameworks CAMEL-AI.org & AutoGen + intro to The $1M AI Agents Challenge - Sign up here meetup.com/ai-agents-lond…

📣 I am happy to announce that Weights & Biases Weave is now integrated with DSPy. 🧶 Weave will automatically capture traces for DSPy. To start tracking, call `weave.init()` and use the library as normal. 👉 Learn more at wandb.github.io/weave/guides/i…

x.com/i/article/1810…

[free ebook] Understanding Deep Learning, 2/07/24 (+68 notebooks) udlbook.github.io/udlbook/

My finetuned models beat OpenAI’s GPT-4 mlops.systems/posts/2024-07-…

Image Self Supervised Learning (SSL) on a Shoestring theadamcolton.github.io/image-ssl-on-a…

Come to our AI Agents London meetup We’ll have a deep dive into two awesome agents frameworks CAMEL-AI.org & AutoGen + intro to The $1M AI Agents Challenge. Sign up here> meetup.com/ai-agents-lond…

⚗️ System 2 prompts: Four techniques to enhance LLM reasoning The "Distilling System 2 into System 1" paper (arxiv.org/html/2407.0602…) looks like Llama 3 Instruct's dataset recipe (note how they only mention Llama 2). The authors build a dataset of complex reasoning tasks with

🐒🐒Tribe AI ✨ Low code tool to rapidly build and coordinate multi-agent teams ✨ Lot's of folks have been asking for a low-code builder for multi-agent LangGraph applications This open source project does exactly that! Code: github.com/StreetLamb/tri…

🏗️Build application specific cognitive architectures, not generic agent infrastructure A custom cognitive architecture will absolutely make your application differentiated, while the generic agent infrastructure is less likely to What's the difference? blog.langchain.dev/why-you-should…

The one-step trap (in AI research) The one-step trap is the common mistake of thinking that all or most of an AI agent’s learned predictions can be one-step ones, with all longer-term predictions generated as needed by iterating the one-step predictions. The most important

Project IDX just got a ton of upgrades. Seriously, a ton. React Native in-browser, AI for comments/tests, database templates, AI agent & Gemini templates. Oh and Android Studio in-browser is coming soon. Here's a rundown.

Llama 3.1 performing multi-step planning, reasoning, and tool calling. This is without an agent framework!

Open. Source. WINS. The results are here: LLaMA 3.1 405b is phenomenal. We now have a SOTA open-source model thanks to AI at Meta . Here's my full test of the model:

Are you building AI Agents? Submit an application to the AI Agents Global Challenge. $1M pool prize, and a chance to get funding from world-class startup investors aiagentschallenge.com

Slides from the AI Agents London meetup - AutoGen AutoGen Multi-Agent Framework drive.google.com/file/d/1v3SC5c…

Slides from the AI Agents London meetup - Camel AI CAMEL-AI.org Collaborative Agents Framework drive.google.com/file/d/1qWDe0l…

Interested in AI Agents? Subscribe free to our newsletter to get the latest research, repos, tools & news on AI Agents agentplex.substack.com/p/agentplex-we…

Thanks Carlos for the invite. Here are the slides from my presentation. Some of my thoughts on - 𝑭𝒊𝒏𝒅𝒊𝒏𝒈 𝒕𝒉𝒆 𝒔𝒄𝒂𝒍𝒊𝒏𝒈 𝒍𝒂𝒘 𝒐𝒇 𝑨𝒈𝒆𝒏𝒕𝒔 at 🐫 CAMEL-AI.org.