Steven Heidel(@stevenheidel) 's Twitter Profileg
Steven Heidel

@stevenheidel

Fine-tuning @OpenAI | ๐Ÿณ๏ธโ€๐ŸŒˆ he/him

ID:2859593312

linkhttps://stevenheidel.com calendar_today17-10-2014 06:24:41

191 Tweets

12,4K Followers

362 Following

Steven Heidel(@stevenheidel) 's Twitter Profile Photo

my twitter replies are a constant battle between team โ€œwhereโ€™s ilyaโ€ and team โ€œwhen gpt-5โ€

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

reminder: when you see headlines about other models beating gpt-4 they're likely talking about gpt-4-0314 from last year, which doesn't have the significant improvements we've added to the latest gpt-4-turbo models

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

delve into the latest gpt-4-turbo model:
- major improvements across the board in our evals (especially math)
- dec 2023 knowledge cutoff

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

people are using ChatGPT to turn bullet points into emails. then on the other end the receiver is using ChatGPT to summarize the email back to bullet points. this could all be avoided if it were just acceptable to send the bullet points in the first place.

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

my favorite feature from today's release is the new side-by-side comparison UI. really fun to see how different models respond to the same prompt. shout-out to Scott Cory and Karolis for launching this! ๐Ÿš€

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

new from fine-tuning:
- weights and biases integration
- saving multiple checkpoints per training job
- a side by side comparison view in the API playground
- and more!

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

happy first sunday after the first full moon that occurs on or after the spring equinox to those who celebrate

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

one of the reasons quantum computers are not yet useful is that qubits lose their state too quickly. an exciting result from my former team shows how we can make this state last longer using โ€œdual-rail qubitsโ€ arxiv.org/abs/2307.08737

account_circle
Steven Heidel(@stevenheidel) 's Twitter Profile Photo

the upside to working at openai is that you know exactly what Greg Brockman is referencing in his tweets

the downside is you still have no idea about roon

account_circle