Choosing the Number of QConfig References per Query for Prompt Augmentation
Ascertain the optimal number of historical query-configuration (QConfig) references to include per query in Booster’s prompt augmentation to maximize recommendation quality while managing context length and inference overhead for large language models.
References
We defer selecting the number of QConfig{s} to enrich the prompt based on each query to future work.
— This is Going to Sound Crazy, But What If We Used Large Language Models to Boost Automatic Database Tuning Algorithms By Leveraging Prior History? We Will Find Better Configurations More Quickly Than Retraining From Scratch!
(2510.17748 - Zhang et al., 20 Oct 2025) in Section 7.7 (Sensitivity Experiments: Input Data)
For a frozen forecaster, what retrieval can offer is the new observations it has never seen and the patterns it has not yet captured; deciding query by query whether a reference can be trusted remains an open problem.
— Retrieval Is Not Enough: Refreshing Memory for Frozen Time-Series Forecasters
(2610.07834 - He et al., 6 Oct 2026) in Conclusion, final paragraph