❤️

I am so excited that xLSTM is out. LSTM is close to my heart - for more than 30 years now. With xLSTM we close the gap to existing state-of-the-art LLMs. With NXAI we have started to build our own European LLMs. I am very proud of my team. arxiv.org/abs/2405.04517

What is the difference between an LLM and an embedding model?

🤖 LLMs: take a string as an input and output a string by predicting the next token.

🤖 Embedding models: take an input and return a list of floats (embeddings), which are the numerical representation of the input.

How LLMs work, clearly explained with visuals:

ScrapeGraphAI: You Only Scrape Once

Neat little web scraping tool powered by LLMs.

LLMs are powerful information extractors so it's not surprising to see the popularity of this Python library and many others.

It works with ollama and other LLM providers.

This is insanely funny!! OAI is proposing extreme security measures to lock down model weights...

On the one hand, several folks are arguing we should open-source LLMs....

On the other hand, secret closed labs like OAI are creating digital walled fortresses!

This totally…

🤖 In an era dominated by big tech, ForumAI & SKALE are joining forces to decentralize AI development!

By leveraging SKALE, ForumAI aims to provide robust, accessible machine learning tools for developers to seamlessly integrate LLMs.

Learn more bit.ly/3wvavET

I've delivered my Advanced Prompting for LLMs training to almost 400 people over the last year.

Here are some of the main themes people learn about:

- applying advanced prompting techniques like chain-of-thought
- best practices to improve the reliability, robustness, and…

#Algorithms on Trees and Graphs (with #Python Code): amzn.to/3IrmKWs
——————
#AI #MachineLearning #BigData #DataScience #RDF #LinkedData #GraphDB #GraphAnalytics #Semantics #KnowledgeGraphs #RAG #LLMs #DataStructures #DataScientists #Coding

'More Agents Is All You Need' - Very interesting Paper.

In this work, multiple LLM agents are used to improve the performance of LLMs. They found that the performance scales with the increase of agents,
using the simple(st) way of sampling and voting.

'When the ensemble size…

it was hard af to build an LLM app that:

- saved chat history
- had many agents + users
- could switch between diff LLMs
- used apps like slack, notion

so we built Julep;
it solved all our problems.

check it out below 👇

Honesty and helpfulness are two central goals of LLMs. But what happens when they are in conflict with one another? 😳

We investigate trade-offs LLMs make, which values they prioritize, and how RLHF and Chain-of-Thought influence these trade-offs: arxiv.org/abs/2402.07282

[1/3]

Esta herramienta es maravillosa. Hay gente que le tiene miedo a los LLMs como Gemini, porque dicen que la IA va a dominarnos, en realidad esto técnicamente no es inteligencia artificial como tal.
Gemini me ha ayudado a entender y a traducir documentos escritos en francés antiguo.

Imagine you are an AI Solutions Provider, offering services like LLMs, prompt libraries, and computational resources to your clients.

xLSTM: Extended Long Short-Term Memory

Attempts to scale LSTMs to billions of parameters using the latest techniques from modern LLMs and mitigating common limitations of LSTMs.

To enable LSTMs the ability to revise storage decisions, they introduce exponential gating and a new…

A superb idea to reduce LLM Refusals. Great for de-censoring instruct models such as Llama3's official instruct finetunes.

Refusal in LLMs is mediated by a single direction 🔥

✨ Why re-finetune the entire model when you can just shut off the part that refuses? ✨

'Refusal is…

Introducing SUQL (Structured and Unstructured Query Language), a novel method to power assistants on hybrid data corpus. It combines SQL relational operators with free text primitives based on retrievers and LLMs, complete with key optimizations

Code available on GitHub and PyPI

xLSTM: Extended Long Short-Term Memory

abs: arxiv.org/abs/2405.04517

Leveraging the latest techniques from modern LLMs, mitigating known limitations of LSTMs (introducing sLSTM and mLSTM memory cells that form the xLSTM blocks), and scaling up results in a highly competitive…

arXiv: Long-context LLMs Struggle with Long In-context Learning

'So the context extension papers started using subprime retrieval evaluations. Gotta keep that citation ring churning. Whenever you hear 'Needle In A Haystack', think 'Shit.''

这个名为《动手学大模型Dive into LLMs》的GitHub项目是一个由上海交通大学扩展的编程实践教程系列。教程专注于教授如何操作大型语言模型（LLMs），涵盖从微调预训练模型到API调用再到知识编辑等多个领域。目标是让大家能够快速掌握并应用这些强大的工具。
GitHub：github.com/Lordog/dive-in…