Being hilariously bad isn’t dangerous. That’s one of the safest ways to be bad. I barely tried Galactica so I don’t have a strong take on how bad it was, but the danger of its out-of-thin-air citations was overblown. Llama2 base is no less hallucinatory and we survived.
LLMS
-
Jais-13B Arabic LLM Launches on Microsoft Azure Cloud
By
–
Jais-13B, the leading Arabic LLM trained on Cerebras CS-2 is now available on the @Azure AI cloud! Very excited for this partnership with @G42ai & @Microsoft
! https://
techcommunity.microsoft.com/t5/ai-machine-
learning-blog/welcoming-mistral-phi-jais-code-llama-nvidia-nemotron-and-more/ba-p/3982699
… -
Faithful dataset representation most valuable in science/engineering, wider audiences need guardrails.
By
–
It’s in sciences and engineering, where users might be expected to understand what a base LLM is and how they’re used, where having faithful representation of the dataset is most valuable. It’s wider audiences that need more guardrails.
-
RL tuning mitigates issues with diversity and creativity cost
By
–
They are! But both are mitigated by RL-based tuning and have similar costs in diversity and creativity. To tune for correctness is to optimize for tokens that, in practice, lead to true output, regardless of how well those choices represent real text. Consider whether an LLM
-
Self-Attention Mechanism in Large Language Models Explained
By
–
Understanding and Coding the Self-Attention Mechanism of Large Language Models From Scratch https://
bit.ly/3YL821l #AI #DeepLearning #MachineLearning #DataScience -
Google delays competing answer to GPT-4 launch
By
–
Looks like we may have to wait even longer for Google’s much-anticipated answer to GPT-4.
-
Custom GPTs will replace 10,000+ AI tools with Recreate Images GPT
By
–
Custom GPTs will replace 10,000+ AI Tools.
— God of Prompt (@godofprompt) 16 novembre 2023
So I built Recreate Images GPT 🖼️
Upload any image and it will recreate it for you!
👉 Try it here: https://t.co/emYyMCF1WK#CustomGPT #GPTs #GPT4 #GPTBuilder #GPTStore #GPT4Vision #ChatGPTplus #ChatGPT4 #Godofprompt #OpenAI… pic.twitter.com/GiCXXaHecWCustom GPTs will replace 10,000+ AI Tools. So I built Recreate Images GPT Upload any image and it will recreate it for you! Try it here: https://
chat.openai.com/g/g-DqViiXYrU-
recreate-images-gpt
… #CustomGPT #GPTs #GPT4 #GPTBuilder #GPTStore #GPT4Vision #ChatGPTplus #ChatGPT4 #Godofprompt #OpenAI -
Yann LeCun defends Galactica: Hallucination isn’t a crime, valuable models need less.
By
–
Yann LeCun is right: Galactica was wronged. Hallucination is not a crime. Models designed to hallucinate less (none eliminate it) hold great value — in dollars, perhaps most. Alignment instills reason and diligence as much as any other of our mores. In many ways, models are more
-
RAG Training Flaw: Models Fail to Ignore Irrelevant Context
By
–
I think they made to common mistake during RAG-training of pretty much *always* having relevant context available, so the model never learns to ignore it.
-
ML Community’s Growing Interest in Reinforcement Learning and RLHF
By
–
And ML people are very interested in RL nowadays, thanks to RLHF's success!