Can an LLM Learn Preferences from Choice Data?
I am happy to announce that we have finished revising our paper on this topic. We study whether LLMs can infer preferences from choice data and then make preference-consistent recommendations in new problems.