Anthropic says Claude solved a nine-loop scattering-amplitude problem in planar N=4 super-Yang-Mills, a simplified model physicists use as a testing ground. Scattering amplitudes help predict particle behavior; “loops” add increasingly precise corrections but require exponentially more computation. The company says the previous record in this model was eight loops. Given a single prompt, Claude ran largely unsupervised for days in Claude Science, using methods developed by SLAC physicist Lance Dixon and colleagues. Anthropic puts the total cost at a few thousand dollars and says Dixon independently verified the result.