Liran Tal on X: "Please don't consider the use of the Function constructor as a safer eval, I beg you 🙏" / X
When people walk into an AAC eval or training the things they say are often “I need to know how to program a button”. “I need to kn... | Instagram
![PDF] PROXYQA: An Alternative Framework for Evaluating Long-Form Text Generation with Large Language Models | Semantic Scholar PDF] PROXYQA: An Alternative Framework for Evaluating Long-Form Text Generation with Large Language Models | Semantic Scholar](https://d3i71xaburhd42.cloudfront.net/d891face4565ef3970c1a0965d8126456651f81e/1-Figure1-1.png)