Agent-500
Emerging1papers using it
2026first seen
Agent-500 is a dataset used to evaluate the ability to identify the source of responses generated by large language models (LLMs) based on varying prompts rather than a fixed input set.
Agent-500 is a dataset used to evaluate the ability to identify the source of responses generated by large language models (LLMs) based on varying prompts rather than a fixed input set.