I am inspired by the notion that chain-of-thought reasoning in language models is a side product of training on code. But I also would like to see more evidence.
Tag: AI
My own personal Žižek
A few days ago, I woke up to the news that Slavoj Žižek wrote an article published on Die Zeit (behind paywall) about my latest work, the Infinite Conversation. I was elated to realize that the project had eventually landed on his desk, and overjoyed to see that he opened his article, in perfect Žižekian …
Stoked to see the Infinite Conversation exhibited at the Audio Art Festival in Kraków.
After googling the answer to a technical question for several minutes, with no good leads, I finally got a correct one-shot answer using GPT-3 using the same query as prompt. It’s happening.
Easy enough even for Dall-e.
Inspiration from: https://inthesetimes.com/article/in-you-more-than-yourself
I agree with the position stated in this post.
haiku barada nikto
Check it out. After working earlier this year with Transformers, I gave GPT-NEO a try for Haiku generation. Not bad! Next step: add a rudder. Also, I think it would be cool to find a visualization method that would display branching and make the generation process interactive/iterative. …
Haiku Generation ~ A Transformer Based Approach ~ With Lots Of Control
Where I teach a GPT model how to generate haiku poems on any topic. …
I’m asking Aleph2Image and TheBigSleep to visualize album names. Here’s the first one. Mustache and colorful uniforms are good hints.
Something nautical, but also what’s with all that yellow?
Love this one. Metal as it should be. That hat is a dead giveaway.
Grungy dog.
Why is OpenAI calling this “erroneous classification”? The model is just doing its Magritte thing, as any good model would.
