people keep saying AI is moving so fast. some days I agree, but some days I'm not sure – so many papers published, but I don't feel like we're making that many fundamental breakthroughs. to cap off 2023, here's a list of things we still don't know about language models: – how
LLMS
-
Language Models Power Compared to Nuclear and Internet Technology
By
–
I don't think language models are "the most powerful technology"; I think that title would go to something else, perhaps nuclear bombs, or battleships, or The Internet
-
LangChain State of AI 2023: Year Review Trends
By
–
As the year draws to a close, highlighting two pieces of content that really summed up 2023 for LangChain: LangChain State of AI 2023: highlighting all the trends in models, vectorstores, retrieval and evaluation that we saw over the past year https://
blog.langchain.dev/langchain-stat
e-of-ai-2023/
… TED -
TinyGPT-V: Efficient Multimodal Large Language Model
By
–
TinyGPT-V: Efficient Multimodal Large Language Model via Small Backbones Zhengqing Yuan, Zhaoxu Li, Lichao Sun : https://
arxiv.org/abs/2312.16862 #Artificialintelligence #ChatGPT #DeepLearning -
FAR AI Researchers Discover Emerging Flaws in GPT-4
By
–
[#Article] Researchers from FAR AI reveal emerging flaws in GPT-4 https://actuia.com/actualite/des-chercheurs-de-far-ai-revelent-les-failles-emergentes-de-gpt-4/
… #AI #artificialintelligence -
4-bit Compression Performance Trade-offs for AI Models
By
–
This is incredible! What is the performance like? Is 4-bit compression too much, even for more generic use cases?
-
Groq LPU Demonstrates Mistral AI 7B Model Performance
By
–
"and other models?" Here's a quick demo of us running @MistralAI 7B, a smaller #LLM, on a Groq LPU™ system. No lag, just "fluid & fluent" experiences, from the (#LPU) Language Processing Unit advantages. http://
Groq.com #groqspeed #betterongroq https://
youtu.be/9c078xKGwdU -
Why AI Alignment Undervalued Compared to Global Priorities
By
–
certainly AI alignment (on a high level) is an important issue but why is it undervalued relative to eg world hunger or pandemic preparedness? and given this, why is studying language models considered a reasonable path towards eventually saving the world?
-
Fine-tuned Open-Source Models Match GPT-4 at Lower Cost
By
–
You don't need an expensive, closed model for most enterprise use cases. Most companies can use a fine-tuned open-source model and match GPT-4 performance for a small fraction of the cost. Here is what you need to know: https://
blog.abacus.ai/blog/2023/10/1
9/open-source-llms-fine-tunes-and-rag-based-vector-store-apis/
… -
Stanford Research Challenges AGI Fears and LLM Capabilities
By
–
Top AI stories of 2023: Stanford researchers showed that fears of AGI are unfounded. Large language models are not greater than the sum of their parts. https://
stanford.io/41IMyVF