Tag
This paper investigates how AI agents can learn explicit social norms from human behavior to improve coordination in dynamic interactions, using pedestrian-vehicle scenarios as a testbed. The proposed norm-informed LLM outperforms baselines and human-human interactions by a significant margin.
NormAct is a benchmark that evaluates embodied planning agents on hidden social norm compliance, revealing that state-of-the-art MLLMs achieve 67.3% goal achievement but only 26.4% norm compliance, and proposes NormPerceptor to improve task success from 24.2% to 46.7%.
LoSoNA is a benchmark for evaluating LLMs' ability to infer and adapt to local social norms in group chat conversations. It tests models on their capacity to pick up hidden norms from precedent and respond appropriately.
This paper proposes a framework for measuring social norms alignment in naturalistic, free-form settings through solution matching, introduces a dataset of 3k social dilemmas in Danish, and evaluates alignment between LLMs and human responses.