“Why the hell did you recommend THIS to me? Your algorithm is trash!”
If you’ve ever shipped a recommender, you know that moment: the metrics look great, but one furious exec can torpedo months of work.
Read the “survival guide”: https://t.co/GnMgX3SgIj
Wizz Air @wizzair sucks. Their website wouldn’t let us check in 3 of 4 passengers on our booking. We had to pay €144 at the airport, and they still refused to refund us even after we provided evidence. If you’re thinking of flying with them, think twice.
“Why the hell did you recommend THIS to me? Your algorithm is trash!”
If you’ve ever shipped a recommender, you know that moment: the metrics look great, but one furious exec can torpedo months of work.
Read the “survival guide”: https://t.co/GnMgX3SgIj
@anshuc@MParakhin Not really interpretable. The authors make an attempt to decode in a very limited case.
As usual, interpretability limits performance. I bet future models will sacrifice the former.
@MParakhin@anshuc BTW, for LLM internal reasoning, discrete vocabulary is not necessary. It can be done in the embedding space. https://t.co/TQfRQG4gWO