Hacker News Reader

Comment by square_usual | Hacker News Reader

Easy enough to explain: they're benchmaxxing. Fable is intelligent but not benchmaxxed. Opus is less intelligent but benchmaxxed.

View on HN

That's a plausible explanation but I'm not seeing evidence for it.

I have a personal benchmark suite of 14 real, non-public tasks. Opus 5 and Fable tied on 10, Opus won on 3, and Fable won on 1. It's a really strong model.

This is an interesting idea, without giving away your benchmarks specifically what kind of stuff do you test? I might try to assemble something myself, it's so hard to determine model quality from the system card these days.

I have actually been planning to open source the framework. Only benchmarks I care about are the ones that look like real work I do. So it makes it easy to trawl your own repos looking for benchmark task candidates from real bug fixes or features. Then as a bonus I can compare the agent’s solution to my own as a reference.

So most look like that but I did include a few one-shot “build an app that solves this problem” and some qualitative design tasks and a tough algorithmic optimization one.

Would open sourcing it make you feel like the results might be compromised? I'd rather keep it private so it's specific to my use cases and not included in any training date (no matter now small the signal in the overall data).

Oh sorry I meant the framework. Though I could see posting a couple of the tasks.

Also working on a product to build tasks from your own work for testing coding agents. Main thing I would offer is to look carefully at the agent trajectories - they love to figure out ways to cheat. Additionally, consider what "winning" means. If just using test pass rate, consider that tests might not encode what good means in your repo. I have been having success using "equivalence with merged PR" as judged by an LLM as a signal.

Yup, trying to be really strategic about testing. I didn't end up sticking with it, but I tried requiring test cases to cite a matching clause in the task assignment. But also: these tests are only indicative. Only a human can score a run.

Yep , same with 5.6. Fable is still the best.

But still nerfed compared to the initial release.

Only when you hit the cyber classifiers.

Honestly that's the simplest explanation and thus likely the correct one.

They don't have a track record of benchmaxxing. The simplest explanation is that the blog post lists the things this model is better than Fable at, but not the things it isn't.

Reply on news.ycombinator.com