3 ms·
Hm. I wonder how well this works versus a large LLM. Seems like something a very strong LLM should be able to handle well with the right prompting. If you can
by ilaksh 2y ago
Hm. I wonder how well this works versus a large LLM. Seems like something a very strong LLM should be able to handle well with the right prompting.
If you can handle it with just phrases that would save a lot of time and money though.
- luke-stanley 2y agoYeah, presumably words with vague meanings, low semantic quality are easy to find. Still, a Bert model for it would be fun.