Search for a command to run...
ProgramBench is a research benchmark for assessing whether language models can generate complete, functional programs from scratch
ProgramBench is a research benchmark for assessing whether language models can generate complete, functional programs from scratch. It provides a comprehensive evaluation framework built in Python to test LLM capabilities in code generation and program synthesis tasks.
Provides a comprehensive evaluation framework built in Python to test LLM capabilities in code generation and program synthesis tasks.
No comments yet. Be the first!
Weighted observed activity · updated Jul 15
567 products launched in 2026 · 100% still active