It’s pretty cool how hallucinations emerged as a mainstream research topic due to ChatGPT’s ability to say “Sorry, I can’t answer that.” In the past (i.e., a few years ago), AI would always guess, so we would simply compute: accuracy = num_correct / num_examples. Now we are
LLMS
-
GPT-4 System Card: Critical Analysis of AGI Claims and AI Pause
By
–
#5: Episode 11: A GPT-4 Fan Fiction Novella, originally recorded on April 7, 2023, in which @alexhanna and I take apart the GPT-4 "system card", the "Sparks of AGI paper" and the #AIPause letter.
-
Stanford CS25 Transformers United V3 New Lectures Released
By
–
[New lectures ] Transformers United – Stanford CS25 The new lectures of Transformers United class V3 have been realized(gradually). Current lectures are about foundational models and generalist agents. Lecture videos: https://
youtube.com/playlist?list=
PLoROMvodv4rNiJRchCzutFw5ItR_Z27CM
…
Website: https://
web.stanford.edu/class/cs25/ -
Multimodal RAG with LangChain and GPT-4 Vision Guide
By
–
Multimodal RAG using Langchain Expression Language And GPT4-Vision Great guide from Plaban Nayak on multimodal RAG. We suspect this will become more and more popular – so now's a great time to become an expert! Blog: https://
medium.aiplanet.com/multimodal-rag
-using-langchain-expression-language-and-gpt4-vision-8a94c8b02d21
… -
Quivr: Local GenAI Second Brain with Llama2 and Docker
By
–
Quivr: your GenAI 2nd Brain One of the most popular RAG apps out their is Quivr. Even better: it can be run completely locally Check out this awesome video by @mesudarshan on how to run it – Llama2 model with @OLLAMA – Uses @Docker & @supabase & @langchain – LangSmith for
-
OpenAI generates $130M monthly from ChatGPT subscriptions and API access
By
–
“OpenAI is generating at least $130 million per month from the sale of subscriptions to ChatGPT and from letting software developers access its models through an application programming interface.”
-
Softmax and Linear Attention Integration in Neural Networks
By
–
Agent Attention: On the Integration of Softmax and Linear Attention Han et al.: https://
arxiv.org/abs/2312.08874 #ArtificialIntelligence #DeepLearning #MachineLearning -
2023 AI Advancements: Music, Robots, GPT-4, and Ethics
By
–
This year, we saw amazing AI advancements with creations like AI music and task-oriented robots and breakthroughs like GPT-4, alongside heated debates on AI ethics. A quick recap of the year in AI
-
Groq LPU Inference Engine Speed Demo with Advanced Prompt Techniques
By
–
A different #prompt technique #demo. Using a larger input #token set on 1st prompt to summarize, and analyze, info not existing when a model was trained. A short 3min vid showing off the inference #groqspeed of our LPU™ Inference Engine. #LLM #AI # https://
youtu.be/Xpm_f4fzWno -
LLM OS: The Future of Local Computing Optimization
By
–
A powerful LLM OS is the only thing capable of fully utilizing your local compute. For the last decade or so most consumer computers were primarily sold on the basis of either their video editing performance (M1, etc) or graphical capabilities. Most users didn't need much else