papersTODAY 04:00 UTC
FriendBench Benchmark Tests Whether AI Can Tell Friends From Strangers
Researchers introduced FriendBench, a benchmark that evaluates how well humans and multimodal large language models can judge whether two people in a short video clip are already acquainted or meeting for the first time. The task uses 20-second recordings of ice-breaker conversations, where cues come from behavior and body language rather than spoken content alone. The work aims to measure social perception abilities that go beyond text-based reasoning.
FriendBenchhuman evaluationmultimodal-llmsnonverbal-communicationsocial-perceptionvideo-understanding
COVERAGE · 2 REPORTS · LINKS GO TO THE ORIGINAL OUTLETS
arXiv cs.AIFriendBench: Benchmarking Dyadic Familiarity Inference in Humans and Multimodal Large Language Models ↗TODAY 04:00 UTC
arXiv cs.CLFriendBench: Benchmarking Dyadic Familiarity Inference in Humans and Multimodal Large Language Models ↗TODAY 04:00 UTC