ThrowBench: Benchmarking LLMs by Predicting Runtime Exceptions

Open in new window