News for hackers. Stuff that matters.
Re:Claude 3.5 Sonnet Reproduces BIG-Bench Canary String by jjcm on Thursday July 11, 2024 @05:10PM (#40940639)
> Even though the latter is the readme, it has the guid, and there's no reason not to ignore every document containing the GUID.

I disagree with this. There are plenty of websites out there that talk about LLM training in general, and have sections dedicated to canary strings. This page for example has the GUID in it: https://ravinkumar.com/GenAiGuidebook/deepdive/BigBench.html

I'd argue that it is something that LLMs should train on. Having context of how LLMs work is something that isn't related to the benchmark data at all. Just because the GUID shows up as an example doesn't mean the benchmark data is present on the page.

Re:Claude 3.5 Sonnet Reproduces BIG-Bench Canary String by PostOnce on Thursday July 11, 2024 @08:04PM (#40941702)
Even so, they can drop from their training set a few dozen pages from the internet and it won't be a huge loss among the trillions of documents.

The LLM will still know how LLMs work without having trained on the handful of documents containing THAT SPECIFIC canary string, because other documents will mention the concept of a canary string without that EXACT GUID.

Better to do that and be on the safe side and look honest than have people believe your company not really competitive.

Anything else is a risk for no gain in an industry theoretically worth trillions.

Re:Claude 3.5 Sonnet Reproduces BIG-Bench Canary String by jhugo on Friday July 12, 2024 @04:27AM (#40943663)
So just to understand this correctly, you're suggesting that they have some people/processes dedicated to keeping track of any canary string published by anyone (or some defined subset of "anyone"), and updating their ingest to ignore any documents that contain those strings?
Re:Claude 3.5 Sonnet Reproduces BIG-Bench Canary String by PostOnce on Friday July 12, 2024 @07:58AM (#40944716)
I explained above that they don't have to care, but if they're going to want to be included on major industry benchmarks like those created by Google, they should probably go to the effort of ignoring the half dozen or so notable benchmarks with canary GUIDs. Blacklisting a GUID is not difficult, not even at web-scale.