CreativityBench: Evaluating Agent Creative Reasoning via Affordance-Based Tool Repurposing
Learn to evaluate agent creative reasoning using CreativityBench, a benchmark for affordance-based tool repurposing, and improve your understanding of AI's problem-solving capabilities
- Build a CreativityBench environment to test agent creative reasoning
- Run experiments using the benchmark to evaluate tool repurposing capabilities
- Configure agents to reason about affordances and attributes of objects
- Test the agents' ability to repurpose tools in novel situations
- Apply the insights from CreativityBench to improve AI model performance on creative problem-solving tasks
AI researchers and developers can benefit from this benchmark to test and improve their models' creative reasoning abilities, while product managers can use it to evaluate the potential of AI-powered tools for creative problem-solving
💡 Creative problem-solving in AI can be evaluated and improved using benchmarks like CreativityBench, which focuses on affordance-based tool repurposing
🤖 Introducing CreativityBench: a benchmark for evaluating agent creative reasoning via affordance-based tool repurposing 🛠️ #AI #Creativity
Key Takeaways
Learn to evaluate agent creative reasoning using CreativityBench, a benchmark for affordance-based tool repurposing, and improve your understanding of AI's problem-solving capabilities
Full Article
Abstract:
arXiv:2605.02910v2 Announce Type: new Abstract: Recent advances in large language models have led to strong performance on reasoning and environment-interaction tasks, yet their ability for creative problem-solving remains underexplored. We study this capability through the lens of creative tool use, where a model repurposes available objects by reasoning about their affordances and attributes rather than relying on canonical usage. As a first step, we introduce CreativityBench, a benchmark for
DeepCamp AI