Decompose, Don't Expand: Capability-Independent AI Systems by Construction
arXiv:2610.38144v1 Announce Type: new Abstract: A recurring design choice in building AI systems is whether, when a task is too hard for a model, to expand the model (make it larger, better-prompted, fine-tuned), or to decompose the task (extract the parts that admit a deterministic solution to a host, leaving the model only the irreducibly fuzzy residue). We make three claims. First, that many tasks commonly delegated whole to a model are in fact deterministic tasks with a small generative core
阅读原始来源
来源记录
研究分析
研究主张
论文标识：2610.38144。 来源描述：arXiv:2610.38144v1 Announce Type: new Abstract: A recurring design choice in building AI systems is whether, when a task is too hard for a model, to expand the model (make it larger, better-prompted, fine-tuned), or to decompose the task (extract the parts that admit a deterministic solution to a host, leaving the model only the irreducibly fuzzy residue). We make three claims. First, that many tasks commonly delegated whole to a model are in fact deterministic tasks with a small generative core
评测与比较记录
可供比较的近期同类记录：“Underdog shows off a 32MB speech model in new demo” (TechCrunch, 2026年11月6日)；“Redacting Personal Data: A Speed and Accuracy Comparison of Regex, Presidio and Hybrid Designs” (arXiv, 2026年11月6日)。主题相近仅用于导航，未核验其是否属于同一需求或独立证据。
结论适用范围
结论只能在所述模型版本、任务、数据与推理预算内解释。社区票数、下载量或榜单位置不能证明另一工作流中的可靠性。
有效性与复现缺口
需核对样本选择、基线配置、不确定性、数据污染风险、负面结果以及代码或权重可获取性；此处没有本站复现实验记录。
复核路线
阅读论文或模型卡及评测附录，针对明确基线复现一个有限结论，再测试代表性用户任务并记录结论失效的场景。