Papers by Yu Ying Chiu
CulturalBench: A Robust, Diverse and Challenging Benchmark for Measuring LMs’ Cultural Knowledge Through Human-AI Red-Teaming (2025.acl-long)
Copied to clipboard
Yu Ying Chiu, Liwei Jiang, Bill Yuchen Lin, Chan Young Park, Shuyue Stella Li, Sahithya Ravi, Mehar Bhatia, Maria Antoniak, Yulia Tsvetkov, Vered Shwartz, Yejin Choi
| Challenge: | CulturalBench is a set of 1,696 human-written and human-verified questions to assess LMs’ cultural knowledge covering 45 global regions including underrepresented ones like Bangladesh, Zimbabwe, and Peru. |
| Approach: | They construct a set of 1,696 human-written and human-verified questions to assess LMs' cultural knowledge, covering 45 global regions including underrepresented ones like Bangladesh, Zimbabwe, and Peru. |
| Outcome: | The proposed model outperforms other models across cultures, while underperforming on questions related to North Africa, South America and Middle East. |
Humanoid Agents: Platform for Simulating Human-like Generative Agents (2023.emnlp-demo)
Copied to clipboard
| Challenge: | Humanoid Agents aims to guide Generative Agents to behave more like humans using System 1 processing . we introduce three elements of System 1 that can influence their behavior . |
| Approach: | They propose a system that guides Generative Agents to behave more like humans . they introduce three elements of System 1 processing that can influence their behavior . humanoid agents can use these dynamic elements to adapt their daily activities and conversations . |
| Outcome: | The proposed system guides Generative Agents to behave more like humans . it incorporates three elements of System 1 processing that can influence their behavior . humanoid agents can adapt their daily activities and conversations with other agents . |