|
| ▲ | frabcus 9 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. |
| |
| ▲ | throwuxiytayq an hour ago | parent [-] | | What if the next token represents a wrong or low-quality answer, but would have only been picked 10% of the time, but now it's picked 20% of the time? Doesn't that obviously decrease the model quality, even though "it might have picked that token anyway"? | | |
| ▲ | dan-robertson 23 minutes ago | parent [-] | | It would be picked 10% of the time with watermarking. The randomness properties of the PRNG will be very similar to other random number generators, it is just chosen to be vulnerable to a particular cryptanalytic attack (that requires a private key known only to anthropic). I think of it like the Dual_EC_DRGB generator rather than a biased coin. |
|
|
|
| ▲ | brokencode 6 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. |
| |
| ▲ | northzen 3 hours ago | parent [-] | | Give me an example how would you watermark a single short sentence like "I like turtles"? | | |
| ▲ | qbit42 2 hours ago | parent [-] | | Watermarking just alters the pseudorandom number generator. If "I like turtles" was previously the response to your prompt with probability 100%, it will still be so. This is why watermarking is only effective for long strings of text |
|
|
|
| ▲ | qgin 7 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. |
