Signal2026-08-03
arXiv

FriendBench: Benchmarking Dyadic Familiarity Inference in Humans and Multimodal Large Language Models

Part of

Prompt-Based Fairness Debiasing

Open primary sourceRead on Substack