SMART is a new four-dimension benchmark for LLM mathematical problem solving that reveals large capability differences across models and introduces an All-Pass Score to measure genuine problem-solving ability.
Three-shot Examples [Examples] The Given Question: [Seed question] Now, analyze the next math problem
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
SMART: Self-Generating and Self-Validating Multi-Dimensional Assessment for LLMs' Mathematical Problem Solving
SMART is a new four-dimension benchmark for LLM mathematical problem solving that reveals large capability differences across models and introduces an All-Pass Score to measure genuine problem-solving ability.