Model5h ago
Claude Matches Human Experts In Bioinformatics
BioMysteryBench is the new evaluation framework proving Claude can solve complex bioinformatics problems. Anthropic claims the model matches human expert performance on these specialized tasks. While the results are promising, the company acknowledges significant caveats regarding the benchmark's scope. Practitioners should view these gains as a strong signal for specialized scientific reasoning.