用梯度数据价值评估优化博弈论运动规划的训练顺序，提升性能并降低波动。
Gradient-Based Data Valuation Improves Curriculum Learning for Game-Theoretic Motion Planning
- 基于梯度相似性为训练场景赋权，构建更优课程学习路径。
- 平均规划误差达1.704米，显著优于基于元数据的课程（1.822米）。
- 适合追求高效训练与鲁棒性的自动驾驶规划研究者使用。
我们证明，基于梯度的数据价值评估能生成显著优于基于元数据启发式方法的课程排序，用于训练博弈论运动规划器。具体而言，我们在nuPlan基准上对GameFormer应用TracIn梯度相似性评分，按场景对验证损失下降的贡献权重构建课程。在三个随机种子下，TracIn加权课程实现平均规划ADE为1.704±0.029米，显著优于基于交互难度的元数据课程（1.822±0.014米；配对t检验p=0.021，Cohen's d_z=3.88），且方差低于均匀基线（1.772±0.134米）。分析显示，TracIn得分与场景元数据几乎正交（Spearman ρ=-0.014），表明梯度方法捕捉到人工特征无法察觉的训练动态。进一步发现，硬性数据选择失效：仅20%的TracIn精选子集性能下降两倍，而相同评分下的全数据课程加权反而取得最佳效果。这些结果确立了梯度数据价值评估在博弈论规划中提升样本效率的实际价值。
原文摘要 · Abstract (English)
We demonstrate that gradient-based data valuation produces curriculum orderings that significantly outperform metadata-based heuristics for training game-theoretic motion planners. Specifically, we apply TracIn gradient-similarity scoring to GameFormer on the nuPlan benchmark and construct a curriculum that weights training scenarios by their estimated contribution to validation loss reduction. Across three random seeds, the TracIn-weighted curriculum achieves a mean planning ADE of $1.704\pm0.029$\,m, significantly outperforming the metadata-based interaction-difficulty curriculum ($1.822\pm0.014$\,m; paired $t$-test $p=0.021$, Cohen's $d_z=3.88$) while exhibiting lower variance than the uniform baseline ($1.772\pm0.134$\,m). Our analysis reveals that TracIn scores and scenario metadata are nearly orthogonal (Spearman $ρ=-0.014$), indicating that gradient-based valuation captures training dynamics invisible to hand-crafted features. We further show that gradient-based curriculum weighting succeeds where hard data selection fails: TracIn-curated 20\% subsets degrade performance by $2\times$, whereas full-data curriculum weighting with the same scores yields the best results. These findings establish gradient-based data valuation as a practical tool for improving sample efficiency in game-theoretic planning.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。