部署预算下的基础模型预测
Prediction with Foundation Models under Deployment Budgets
Zhaohui Xu · Xiaofei Xu · Doudou Zhou
中文摘要
基础模型可为下游任务提供基于大规模预训练知识的预测,但不同模型的任务特定准确率和查询成本各异。该框架在每输入平均查询成本预算约束下改进任务预测,利用训练数据学习查询哪些模型以及如何将输出与无需查询即可获得的特征结合。第一种方法对每个输入查询相同模型;第二种方法允许查询模型随输入变化,并用Q-learning决定下一个查询哪个模型以及何时停止。对于第一种方法,证明多个单独不准确的输出可联合超越最佳单一输出;当回归函数仅通过预算内模型依赖输出时达到极小极大率,否则预算内任何方法都可能被迫使达到更慢的率,而该方法能达到该更慢率。对于第二种方法,证明何时随输入变化查询模型严格优于对每个输入查询相同模型,并给出有限样本下的预测风险和平均成本界。在模拟实验和多模态笔记本电脑价格应用中,两种方法均优于现有方法。
关键要点
- 01问题:基础模型任务准确率与查询成本不同,需在平均查询成本预算下优化预测。
- 02方法一:对每个输入查询相同模型,学习组合输出与免费特征。
- 03方法二:用Q-learning动态选择查询模型及停止时机。
- 04结果:多模型联合可超越最佳单模型,并达到极小极大率或更慢率。
- 05实验:模拟与多模态笔记本价格应用中均优于现有方法。
解读
尚无解读。
原始英文摘要
arXiv:2611.12466v2 Announce Type: replace Abstract: Foundation models can supply predictions for a downstream task that draw on knowledge from extensive pretraining, but they differ in task-specific accuracy and query cost. We develop a framework that improves prediction for the task under a budget on the average query cost per input, using training data to learn which models to query and how to combine their outputs with features available without querying. The first method queries the same models for every input; the second lets them vary across inputs and uses Q-learning to decide which model to query next and when to stop. For the first method, we show that several individually inaccurate outputs can jointly outperform the best single one. It attains the minimax rate when the regression function depends on the outputs only through models within the budget. Otherwise, any method within the budget can be forced to a slower rate, which the first method attains. For the second method, we show when varying the queried models across inputs strictly improves on querying the same models for every input, and bound its prediction risk and average cost in finite samples. In simulations and a multimodal laptop price application, both methods outperform existing methods.