AgentCAT is a multi-agent LLM framework that simulates examinee responses, question selection via knowledge graphs, and supervision to benchmark dynamic CAT on real-world datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
AgentCAT: Simulating Computerized Adaptive Testing via Multi-Agent Large Language Models
AgentCAT is a multi-agent LLM framework that simulates examinee responses, question selection via knowledge graphs, and supervision to benchmark dynamic CAT on real-world datasets.