BESPOKE

BESPOKE is a benchmark designed to evaluate personalization in search-augmented large language models by collecting authentic chat and search histories from humans, along with fine-grained preference scores and feedback.

Papers using BESPOKE (2)

BESPOKE: Benchmark for Search-Augmented Large Language Model Personalization via Diagnostic Feedback2025BESPOKE: Benchmark for Search-Augmented Large Language Model Personalization via Diagnostic Feedback2025