Real Human
Research on "Real Human" focuses on understanding and replicating human capabilities, particularly in perception, cognition, and social interaction, using artificial intelligence models. Current efforts concentrate on developing and evaluating large language models (LLMs) and large vision-language models (LVLMs), often incorporating architectures like transformers and diffusion models, to benchmark AI performance against human benchmarks in tasks ranging from visual perception and emotion recognition to complex decision-making and social interaction. These studies aim to improve AI systems' alignment with human behavior and understanding, ultimately impacting fields like human-computer interaction, robotics, and social sciences.
Papers
Characterizing Similarities and Divergences in Conversational Tones in Humans and LLMs by Sampling with People
Dun-Ming Huang, Pol Van Rijn, Ilia Sucholutsky, Raja Marjieh, Nori Jacoby
mCSQA: Multilingual Commonsense Reasoning Dataset with Unified Creation Strategy by Language Models and Humans
Yusuke Sakai, Hidetaka Kamigaito, Taro Watanabe
The 3D-PC: a benchmark for visual perspective taking in humans and machines
Drew Linsley, Peisen Zhou, Alekh Karkada Ashok, Akash Nagaraj, Gaurav Gaonkar, Francis E Lewis, Zygmunt Pizlo, Thomas Serre