用稀疏自编码器揭开端到端自动驾驶模型的内部结构
Mechanistic Interpretability for End-to-End Self-Driving
有人把可解释性技术用到了端到端自驾模型上，能读出红绿灯、刹车这些内部特征，还能改潜变量控制车辆行为，代码也开源了。
研究团队在单摄像头端到端路径规划模型的内部激活上训练 Sparse Autoencoders（SAE），无需监督即可从 SAE 潜变量中恢复路线意图，准确率达 98.3%。通过成对反事实图像编辑，论文显示特定 SAE 特征与红绿灯状态、停车标志等场景变化存在系统性对应。对单个 SAE 潜变量进行引导（steering），会让规划器输出中的横向位移、曲率和刹车行为产生可预测的单调变化。基于 SAE 潜变量训练的轻量适配器将轨迹选择误差降低 3.44%，优于不含 SAE 潜变量的同规格适配器。代码与可视化已在项目页面开源。
Mechanistic Interpretability for End-to-End Self-Driving
End-to-end (E2E) autonomous driving models achieve strong planning performance but remain largely opaque, limiting observability and trust. We apply mechanistic interpretability to E2E driving by training Sparse Autoencoders (SAEs) on the internal activations of a simple, single-camera end-to-end path planning model. Unsupervised clustering of SAE latents reveals interpretable structure aligned with route intent, vehicle speed, and braking behavior, recovering route intent with 98.3% accuracy without supervision. Through paired counterfactual image edits, we further show that specific SAE features are visually grounded and systematically respond to scene changes such as traffic light state and stop sign presence. Steering individual SAE latents produces predictable, monotonic changes in planner outputs including lateral displacement, curvature, and braking. Finally, we demonstrate a practical application of these representations: we build a lightweight adapter trained on SAE latents that reduces trajectory selection error by 3.44%, outperforming an equivalent adapter that omits SAE latents, confirming that the sparse representation captures useful information for interpretability. Code and visualizations are available at https://mgagvani.github.io/mechanistic-interpretability-for-e2e-driving/