RAG From Scratch: Query Translation (Step-Back) Our RAG From Scratch video series walks through impt RAG concepts in short / focused videos w/ code. This is the fourth in our videos on Query Translation, focused on step-back prompting from @denny_zhou
's group at DeepMind.
LLMS
-
RAG Query Translation Step-Back Prompting Technique
By
–
-
AI Chatbots with Current Information: Beyond Search Engines
By
–
AI Chatbots with updated current information that can answer questions, provide opinions on current events, and help with problem-solving are very useful. Google Search is still irreplaceable even with ChatGPT/Perplexity/Bard but imagine an application that is both a chatbot and
-
Magic Startup Achieves AI Reasoning Breakthrough for Engineer Agents
By
–
Magic, a startup building AI software engineer coworkers just had a major breakthrough. Per The Information, Magic has claimed to have made a technical breakthrough that could enable “active reasoning” capabilities similar to Q* by OpenAI. They can also process up to 3.5
-
Dolma: Open Corpus Three Trillion Tokens Language Model Pretraining
By
–
Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research Soldaini et al.: https://
arxiv.org/abs/2402.00159 #ArtificialIntelligence #DeepLearning #MachineLearning -
Google Gemma Launch Partner Delivers Optimized LLM Desktop Models
By
–
Announced today, we are collaborating as a launch partner with @Google in delivering Gemma, an optimized series of models that gives users the ability to develop with #LLMs using only a desktop #RTX GPU.
-
Google Gemma Now Supported in AutoTrain No-Code Tool
By
–
Google announced gemma today, and yes, its already supported in your favorite no-code finetuning tool: AutoTrain
-
Language Agent Tree Search: Balancing Exploration and Reward
By
–
3/ Language Agent Tree Search (LATS) Algorithm combines reflection and Monte-Carlo tree search to find the “best” trajectory for a task. The search process balances exploration and expected reward. It has 4 main steps: 1. Select: pick the best next actions based on the
-
LLM Reflection Architecture: Generator and Reflector Pattern
By
–
1/ Basic Reflection Two LLM nodes: a generator and a reflector. The generator tries to answer the request, while the reflector offers constructive criticism. Python: https://
github.com/langchain-ai/l
anggraph/blob/main/examples/reflection/reflection.ipynb
…
Youtube: https://
youtube.com/watch?v=v5ymBT
XNqtk&t=48s
… -
Cascade Speculative Drafting Accelerates LLM Inference Speed
By
–
Cascade Speculative Drafting for Even Faster LLM Inference Chen et al.: https://
arxiv.org/abs/2312.11462 #ArtificialIntelligence #DeepLearning #MachineLearning -
Google Launches Gemma Open Source Models Built on Gemini Technology
By
–
It certainly has been a fun year @Google
: enjoy playing with our open source models Gemma, built from the same research and technology used to create the Gemini models. Blog: https://
blog.google/technology/dev
elopers/gemma-open-models/
…
Tech report: https://
storage.googleapis.com/deepmind-media
/gemma/gemma-report.pdf
…