Matt Turk, Brandon L. Neel, Diana M. Lopez, Zachary D. Smith, Grace N. Johnecheck, James B. Golden · 2026 · arXiv
DOI: 10.48550/arXiv.2610.26829 · open-access PDF
Os agentes de fronteira recuperam bem os metadados estruturais e calculam a geometria das coordenadas de forma menos confiável. Perguntamos quanto de uma pontuação de referência reflete essa lacuna e quanto reflete a configuração de avaliação em torno do modelo. Estudamos o ProteinGeoBench, um conjunto de 131 tarefas de estrutura de proteínas, usando 4.249 lançamentos de quatro modelos de fronteira e 4.486 testes adicionais em 26 modelos. Três dos quatro modelos de fronteira caem dentro de 0,031 um do outro (0,785 a 0,816) sem nenhuma diferença entre eles resolvida, enquanto o quarto pontua 0,707 e todas as três lacunas a ele excluem zero.
Machine translation (Meta M2M100); the English original follows.
Frontier agents retrieve structural metadata well and compute geometry from coordinates less reliably. We ask how much of a benchmark score reflects that gap and how much reflects the evaluation configuration around the model. We study ProteinGeoBench, a suite of 131 protein-structure tasks, using 4,249 rollouts from four frontier models and 4,486 further trials across 26 models. Three of the four frontier models fall within 0.031 of one another (0.785 to 0.816) with no difference between them resolved, while the fourth scores 0.707 and all three gaps to it exclude zero. Three of the four do b
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26829, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange