Papers by Arjun Arunasalam
Implicit Values Embedded in How Humans and LLMs Complete Subjective Everyday Tasks (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large language models (LLMs) can underpin AI assistants that help users with everyday tasks, such as making recommendations or performing basic computation. |
| Approach: | They audit how six popular large language models (LLMs) complete 30 everyday tasks and compare them to 100 human crowdworkers from the US. |
| Outcome: | The LLMs perform 30 tasks and are compared to 100 human crowdworkers in the US. |