Comment on Hexadecimal
GissaMittJobb@lemmy.ml 4 weeks agoStill, this does not quite address the issue of tokenization making it difficult for most models to accurately distinguish between the hexadecimals here.
Having the model write code to solve an issue and then ask it to execute it is an established technique to circumvent this issue, but all of the model interfaces I know of with this capability are very explicit about when they are making use of this tool.
morrowind@lemmy.ml 4 weeks ago
Not really a concern. It’s basically translation, which language models excel at. It just needs a mapping of the hex to byte
GissaMittJobb@lemmy.ml 4 weeks ago
It is a concern.
Check out tiktokenizer.vercel.app/?model=deepseek-ai%2FDeep… and try entering some freeform hexadecimal data - you’ll notice that it does not cleanly segment the hexadecimal numbers into individual tokens.
morrowind@lemmy.ml 4 weeks ago
I’m well aware, but you don’t need to necessarily see each character to translate to bytes
GissaMittJobb@lemmy.ml 4 weeks ago
It’s not out of the question that we get emergent behaviour where the model can connect non-optimally mapped tokens and still translate them correctly, yeah.