"It’s impossible to fine-tune effectively without an eval system which can lead to writing off fine-tuning if you haven't completed this prerequisite." That sounds spot-on to me. Eval systems still feel like a dark art, which would explain why I don't find fine-tuning attractive
RESEARCH
-
LLM as Global Mind: Hegel’s Spirit in AI
By
–
Hegel used the term to describe the universe looking at itself, through the spirit of human culture. The LLM is quite literally a global mind, distilled from the entirity of shared written cultural production if humanity. The prompt constrains it into a single contextual identity
-
The History of Artificial Intelligence and Big Data
By
–
The History of #AI
by @ParmidaBeigi #BigData #ArtificialIntelligence #ML #MI #DataScience cc: @pbalakrishnarao @ravikikan @kuriharan -
GPT-4 Problem Solving: LLMs Match Human Creativity
By
–
AI as problem solver: A test of LLMs on "MacGyver-like" problems requiring novel solutions Out-of-the-box, GPT-4 only does okay, but when prompted to "think" conveniently & divergently, it is close to the average human, and can exceed them in many cases. https://
arxiv.org/pdf/2311.09682
.pdf
… -
10 Free NVIDIA AI Courses to Unlock Your Future
By
–
https://
drdebashisdutta.com/unlock-the-fut
ure-10-must-take-free-ai-courses-from-nvidia/
… -
Fine-Grained Human Feedback Training with Density Diversity Reward Functions
By
–
#MosiacAI researchers and @UW wanted to determine a better approach to Fine-Grained Human Feedback. See how they researched an approach that utilized training and learning from two different reward functions: density and diversity. https://
dbricks.co/49HyKOj -
3 Types of Machine Learning Every Data Scientist Should Know
By
–
3 types of #MachineLearning that every #DataScientist should know by @mdancho84 #AI #ArtificialIntelligence #DataScience #Tech #Technology #Data cc: @bigdata @ravikikan @pbalakrishnarao
-
Artificial Intelligence vs Human Intelligence: Key Differences Explained
By
–
#ArtificialIntelligence vs. human intelligence: Differences explained
by Michael Bennett @TechTargetNews Learn more: https://
buff.ly/4au4cj7 #ArtificialIntelligence #MI #MachineLearning #Tech cc: @ravikikan @yvesmulkers @kuriharan -
AI Model Capabilities and Public Release Strategy Debate
By
–
That’s just…not accurate 1) Could be that it’s more default and more annoying/cumbersome to prompt it out, just like chatgpt is quite wordy. Not saying it’s true, but it’s possible. 2) It’s their main explanation for no near-term public release and literally looks less
-
Adversarial Patches Can Fool Vision-Guided Missiles
By
–
Adversarial attacks can fool vision-guided missiles. Matroid can fool them with a patch.