BigCodeBench shows LLMs achieve at most 60% on 1,140 tasks needing diverse function calls and complex instructions, compared to 97% human performance.
- Example: Before: ‘# TODO: Implement this‘ After: Actual implementation of the required logic
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions
BigCodeBench shows LLMs achieve at most 60% on 1,140 tasks needing diverse function calls and complex instructions, compared to 97% human performance.