Evidence receipt / belief
Published · transcript-backedAlistair Pullen: belief
22 Aug 2024 Latent Space Is finetuning GPT4o worth it? — with Alistair Pullen, Cosine (Genie)
“We believe that if you want a model to behave like a software engineer, it has to be shown how a human software engineer works.”
Source trail
Everything needed to verify it.
- Speaker
- Alistair Pullen
- Attribution
- Verified speaker
- Claim type
- belief
- Recorded
- 22 Aug 2024
- Publisher
- Latent Space
Transcript context
…Welcome back. This is Charlie, your AI cohost. As AI engineers, we have a special focus on coding agents, fine tuning, and synthetic data. And this week, it all comes together with the launch of Cosign's Genie, which reached 50 percent on SWE Bench Lite, 30 percent on the full SWE Bench, and 44 percent on OpenAI's new SWE Bench Verified. All state of the art results by the widest ever margin recorded compared to former leaders Amazon Q and US Autocode Rover. And Factory Code Droid. As a reminder, Cognition Devon went viral with a 14 percent score just five months ago. Cosign did this by working closely with OpenAI to fine tune GPT 4. 0, now generally available to you and me, on billions of tokens of code, much of which was synthetically generated. Hi, I'm Ali. Co founder and CEO of Cosign, a human reasoning lab. And I'd like to show you Genie, our state of the art, fully autonomous software engineering colleague. Genie has the highest score on SWBench in the world. And the way we achieved this was by taking a completely different approach. We believe that if you want a model to behave like a software engineer, it has to be shown how a human software engineer works. We've designed new techniques to derive human reasoning from real examples of software engineers doing their jobs. Our data represents perfect information lineage, incremental knowledge discovery, and step by step decision making. Representing everything a human engineer does logically. By actually training Genie on this unique dataset, rather than simply prompting base models, which is what everyone else is doing, we've seen that we're no longer simply generating random code until some works. It's tackling problems like a human. Alistair Pullen is CEO and co founder of Kozen, and we managed to snag him on a brief trip stateside for a special conversation on building the world's current number one coding agent. Watch out and take care.…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.