1). Learning to Reason with LLMs – a new family of LLMs trained with reinforcement learning to reason before it responds to complex tasks; it produces a long internal chain of thought and exceeds in science, code, and math-related tasks…
LLMS
-
Top ML Papers of the Week: LLMs, Agents, and Research
By
–
The Top ML Papers of the Week (Sep 9 – Sep 15): – Chai-1
– DataGemma
– LLaMa-Omni
– Agent Workflow Memory
– Learning to Reason with LLMs
– Can LLMs Generation Novel Research Ideas Read on for more: -
Evaluating Perplexity Pro search performance against existing AI models
By
–
I am wondering if perplexity pro search on top of existing models would make any difference to this chart
-
o1 and Next-Gen Models Post-Train on IQ Tests via RLHF
By
–
The best way to parse this post is o1 and the next set of models from everyone else have and will post-train (RLHF) on IQ tests.
-
Using Google AI Studio model comparison features
By
–
No worries. Actually AI studio got comparison mode between models. I forgot to post it
-
Groq API Integration with OpenWebUI Tutorial Video
By
–
Check out this video from @mesudarshan (Data Science Basics on YT) for a tutorial on integrating the Groq API (including our newly introduce Llava 1.57B model) with the OpenWebUI.
-
Self-Attention in LLMs Visualized as Directed Graph
By
–
Self-attention in LLMs represented as a directed graph!
-
Perplexity AI entity recognition behavior
By
–
Love it It is getting interesting when there are multiple people with the same name. Pplx thinks it is one person
-
Creating a custom GPT to manage AI memory and audio recording
By
–
I now also willing to have a quick way to undo memory. When I ask it to record an audio for me on mobile it is now always saving it And I don’t want to disable it completely. Had to make a custom GPT for this purpose
-
Comparing OpenAI o1 to GPT-3.5 and future outlook
By
–
Feels like o1 is an analogy of gpt 3.5 tbh, 2025 will be interesting