CreativityBench shows that current LLMs often pick plausible objects for creative tasks but fail to identify the right parts, affordances, and physical mechanisms, with scaling and chain-of-thought providing limited help.
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
-
CreativityBench: Evaluating Agent Creative Reasoning via Affordance-Based Tool Repurposing
CreativityBench shows that current LLMs often pick plausible objects for creative tasks but fail to identify the right parts, affordances, and physical mechanisms, with scaling and chain-of-thought providing limited help.