Top ML Papers of the Week (Nov 20 – Nov 26): – Mirasol3B
– System 2 Attention
– Parallel Speculative Sampling
– Advancing Long-Context LLMs
– Teaching Small LMs To Reason
– LLMs as Collaborators for Medical Reasoning
…
PROMPT ENGINEERING
-
Top ML Papers: Mirasol3B, System 2 Attention, Speculative Sampling
By
–
-
Tool Use Methodology: Retrieval Approach in AI Systems
By
–
There’s a paper describing this approach, for one particular tool (retrieval); idea is the same however for other tools https://
arxiv.org/abs/2310.11511 -
Master ChatGPT Prompt Crafting: AI Dialogue Art Guide
By
–
Dive into the art of #ChatGPT Prompt Crafting with this insightful infographic by GPT AI! Master the dialogue with AI by understanding roles, tasks, formats, and more. Elevate your tech game! Follow @ingliguori for a journey through #DigitalTransformation and #AI insights.
-
Using DALL-E 3 to generate character illustrations with specific prompts
By
–
For this exact video I 1st generated an initial image with DALLE-3 using this prompt: Illustration by Jamie Hewlett, futuristic robot in front of 6 screens, punk caricature characters, bold lines, vibrant colors, dramatic shading. The result looks like this 👀 pic.twitter.com/JOKagRJ6oU
— 🚨 AI News | TestingCatalog (@testingcatalog) 26 novembre 2023For this exact video I 1st generated an initial image with DALLE-3 using this prompt: Illustration by Jamie Hewlett, futuristic robot in front of 6 screens, punk caricature characters, bold lines, vibrant colors, dramatic shading. The result looks like this
-
GPT-4 Vision Applications for Streamlit Apps
By
–
@JefferyTatsuya just dropped an impressive Twitter on my latest #GPT4Vision blog post! Heartfelt thanks to you, Jeffrey! I feel truly honored! (link to the post https://
blog.streamlit.io/7-ways-gpt-4-v
ision-can-uplevel-your-streamlit-apps/
…) -
Distinguishing Humans from LLMs with Chain-of-Thought
By
–
I literally actually don't know what you mean by this. What in this sense distinguishes a human from an LLM using chain-of-thought?
-
Copilot Framework: Human Agency Over AI Automation
By
–
Satya Nadella has the right framework for AI. “Copilot”. Not “Autopilot”. Human agency is still at the premium. You get to decide what you want to do still. You’re just a lot more well informed and have access to the best expert on the topic at the cheapest cost possible.
-
Shorter posts on X, skepticism for ChatGPT advice
By
–
Shorter is better on X — but yes, take everything I post on ChatGPT with a grain of salt for that reason. I used to be careful about posting generation params and preferring temp=0 in the GPT-3 days but with ChatGPT I don’t fight the quicksand.
-
ChatGPT mobile’s system prompt biases towards shorter answers, favoring API models for research.
By
–
ChatGPT on mobile includes a system prompt that asks for shorter answers, included whether or not you use custom instructions. Details like this (also, unannounced updates to the prompt) are why for research data you usually want the API models instead of ChatGPT.
-
ChatGPT Mobile System Prompt: Shorter Answers, API Models for Research Data
By
–
ChatGPT on mobile includes a system prompt that asks for shorter answers, included whether or not you use custom instructions. Details like this (also, unannounced updates to the prompt) are why for research data you usually want the API models instead of ChatGPT.