upvote

points

by flaghacker9 hours ago |
comments
upvoteby pkulak9 hours ago|
next
[-]
> Text watermarking has no effect on output quality

It has an effect, and it's negative. It's hoped that the effect is negligible, and it probably is, but the whole point is that it has an effect.

reply
upvoteby markasoftware7 hours ago|
parent|
next
[-]
Its essentially swapping out the psuedo random number generated with a differently seeded one iirc.

It has an effect on the output, but not the output quality

reply
upvoteby reasonableklout8 hours ago|
parent|
prev|
next
[-]
It seems fine to me. The model is still solving my problems and writing code that works as well as any other.

Google has been watermarking text with SynthID for a while now and nobody complained about it. Why all the fuss about Claude?

It feels like the real reason behind most complaints is that people want to use AI for writing and not have others find out?

reply
upvoteby k1m4 hours ago|
parent|
prev|
next
[-]
It reduces diversity, which they don't talk about much. Wrote about it here. https://blog.keyvan.net/p/ai-text-watermarking-and-quality
reply
upvoteby geysersam5 hours ago|
parent|
prev|
next
[-]
It has an effect, 50% of the time it's positive, increasing the quality of the answer, and 50% of the time it's negative.
reply
upvoteby qgin4 hours ago|
parent|
[-]
These models are already running with a temperature that is >0. This just replaces a random number with a slightly less random number.
reply
upvoteby NewsaHackO7 hours ago|
parent|
prev|
next
[-]
I am pretty sure they did A/B testing to show it didn't. I could gave sworn they even released a quiz were the user has to try and guess which answer is watermarked or not and it was impossible to tell.
reply
upvoteby sejje6 hours ago|
parent|
[-]
That's not the metric, though.

You either output the best version, or you output something else.

You can't do both.

reply
upvoteby frabcus6 hours ago|
parent|
next
[-]
That's not the case, because LLMs are non-deterministic.

It only alters outputs when the last layer of the neural network give significant weights to multiple tokens, and it would anyway have picked a random answer.

Instead it picks a non-random one, but non-random in such a way that you can't tell without the private key of the watermarking.

This mostly adds randomness these days for branches in syntax that make no difference, and the model has no reason to believe make a difference. Anything that matters, it is much more confident in the last layer of weights on the token to use.

reply
upvoteby NewsaHackO6 hours ago|
parent|
prev|
next
[-]
I feel as though you are overlooking simple statistics/confidence intervals. It absolutely possible for two different works to be not have a distinguishable difference in quality.
reply
upvoteby sejje5 hours ago|
parent|
[-]
Agreed, but not if one of them was altered to contain a secret message. That one will have a distinguishable difference in quality. Maybe (almost certainly) negligible, but still there.
reply
upvoteby joshuamorton5 hours ago|
parent|
[-]
This...isn't true.

Others have already said this, but the watermarking is something like "when the model flips a coin picking between two values, always choose heads". It was already flipping a coin. You're not choosing a less good result, you're just using a deterministic process when it was stochastic before.

This will have some impact on outputs, but unless you have some reason to believe that always picking tails was better than always picking heads (in which case, you should be working at one of these companies in model training!) it won't have any impact on output quality.

reply
upvoteby itemize12350 minutes ago|
parent|
prev|
[-]
that's fair. but it outputted non-best version in the first place sometimes too. Now it still output non-best version sometimes.
reply
upvoteby arrrg8 hours ago|
parent|
prev|
[-]
Why do you claim that?

There is no reason why there has to be a negative effect of text watermarking.

reply
upvoteby pkulak7 hours ago|
parent|
[-]
It literally re-weights the output tokens from what the LLM would otherwise have chosen. It _has_ to. It can't be positive, because then that's not watermarking, it's a better LLM.
reply
upvoteby frabcus6 hours ago|
parent|
next
[-]
It's a very unintuitive algorithm, and is pretty clever.

I recommend reading up on it: https://www.nature.com/articles/s41586-024-08025-4

But no, it only ever picks tokens that are in the probability distribution of the last layer, and it might have picked anyway.

reply
upvoteby brokencode4 hours ago|
parent|
prev|
next
[-]
Unless you’re at 0 temperature, there is no single token it would have chosen. It’s always picking one of multiple randomly according to a probability distribution.
reply
upvoteby northzen1 hours ago|
parent|
[-]
Give me an example how would you watermark a single short sentence like "I like turtles"?
reply
upvoteby qgin4 hours ago|
parent|
prev|
[-]
Unless you’re running at temperature 0, there’s not one single token that the model definitely would have chosen each time.
reply
upvoteby exabrial8 hours ago|
prev|
[-]
This is hilarious this keeps being repeated by the true believers ad nauseam.

Also, don't apply EU law to the world. It's a knee jerk reactionary regulation by a bunch of aging ding dongs that can't print their emails.

reply
upvoteby frabcus6 hours ago|
parent|
[-]
You're on Hacker News - I suggest you have technical curiosity and actually understand this very unusual and innovative algorithm, before you claim things about it that aren't true.
reply