Robuta

https://arxiv.org/abs/2503.03750v3 [2503.03750v3] The MASK Benchmark: Disentangling Honesty From Accuracy in AI Systems Abstract page for arXiv paper 2503.03750v3: The MASK Benchmark: Disentangling Honesty From Accuracy in AI Systems accuracy in ai the mask https://arxiv.org/html/2503.03750v3 The MASK Benchmark: Disentangling Honesty From Accuracy in AI Systems accuracy in ai the mask benchmark disentangling honesty