1 paper
Prateek Rajput, Abdoul Aziz Bonkoungou, Yewei Song +4
Current evaluations of LLMs for code generation emphasize functional correctness, overlooking the fact that functionally correct solutions can differ significantly in algorithmic c…