AI - AI Dynamics

→ View original post on X — @jeande_d

Survey on Transformers Applications in Reinforcement Learning

By

@jeande_d

–

12 February 2023 22h21

A Survey on Transformers in Reinforcement Learning Transformers have been the prime mover of almost all modalities, from vision, speech, text and are showing potential applications in RL. This survey paper provides a nice taxonomy of Transformers in RL. https://
arxiv.org/abs/2301.03044

→ View original post on X — @dotcsv

Direct Tool Demonstration: User Feedback and Impressions

By

@dotcsv

–

12 February 2023 21h16

Directo finalizado. Realmente increíble ¿Con qué opinión os quedáis de la herramienta una vez habéis visto cómo funciona?

LLM Criticism and Mathematical Understanding Behind Models

By

–

12 February 2023 20h39

I find the criticism of his piece absurd, but it was hardly meant for people who “understand the matrix math” behind a LLM.

Explaining GPT limitations rationally against hype

By

–

12 February 2023 20h34

That just says it is wrong but doesn’t offer a rationale. I think it is quite a useful way to explain what gtp does and why it is limited. And that is really needed when there is so much hyperbole and over-extrapolation.

→ View original post on X — @aibreakfast

DALL-E 2 creates hyper-realistic police sketches

By

@aibreakfast

–

12 February 2023 20h15

“Hyper-realistic” police sketches of a suspect based on user inputs.

Two developers have used DALL-E 2 to create a forensic sketch program for law enforcement.

Project: https://t.co/J8ufPrxptD pic.twitter.com/poWWVA75Js
— AI Breakfast (@AiBreakfast) 12 février 2023

“Hyper-realistic” police sketches of a suspect based on user inputs. Two developers have used DALL-E 2 to create a forensic sketch program for law enforcement. Project: https://
lablab.ai/event/openai-w
hisper-gpt3-codex-dalle2-hackathon/eagleai/forensic-sketch-airtist
…

→ View original post on X — @karpathy

Content-Conditioned Q&A Assistants: A Future Feature

By

@karpathy

–

12 February 2023 20h13

content-conditioned Q&A assistant is a prominent feature of the future.

Ted Chiang analyzes ChatGPT’s underlying mechanisms and implications

By

–

12 February 2023 20h07

Ted Chiang, a brilliant scifi writer, wrote a brilliant piece on ChatGPT. Well worth reading if you want to understand what’s really going on behind the prompt. https://
newyorker.com/tech/annals-of
-technology/chatgpt-is-a-blurry-jpeg-of-the-web/amp
…

→ View original post on X — @karpathy

Step-by-Step Thinking Needs Correct Answers in GPT Models

By

@karpathy

–

12 February 2023 20h04

One of my favorite results in 2022 was that it's not enough to just think step by step. You must also make sure to get the right answer 😀 https://
sites.google.com/view/automatic
-prompt-engineer
…
(actually a nice insight into a psychology of a GPT; it pays to condition on a high reward)

Meeting in San Francisco to discuss generative AI and ChatGPT

By

–

12 February 2023 20h02

I’m in San Francisco for a few days. Hit me up if you’d like to meet and talk about generative AI, cerebral valley, and the profound brilliance & utter stupidity of chatgpt.