# On the Sample Complexity of End-to-end Training vs. Semantic Abstraction Training

arXiv: https://arxiv.org/abs/1604.06915 · HTML: https://mlsanity.com/paper/1604.06915/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 Apr 2016.

## Built on by (most cited first, 9 of 9)

- [The pitfalls of next-token prediction](https://mlsanity.com/paper/2403.06963/) (2024)
- [From Foresight to Forethought: VLM-In-the-Loop Policy Steering via Latent Alignment](https://mlsanity.com/paper/2502.01828/) (2025)
- [Sub-Task Decomposition Enables Learning in Sequence to Sequence Tasks](https://mlsanity.com/paper/2204.02892/) (2022)
- [On The Specialization of Neural Modules](https://mlsanity.com/paper/2409.14981/) (2024)
- [Learning to Drive Anywhere](https://mlsanity.com/paper/2309.12295/) (2023)
- [Safe Imitation Learning on Real-Life Highway Data for Human-like Autonomous Driving](https://mlsanity.com/paper/2110.04052/) (2021)
- [A Contextual Bandit Approach for Learning to Plan in Environments with Probabilistic Goal Configurations](https://mlsanity.com/paper/2211.16309/) (2022)
- [STEAM & MoSAFE: SOTIF Error-and-Failure Model & Analysis for AI-Enabled Driving Automation](https://mlsanity.com/paper/2312.09559/) (2023)
- [LASMP: Language Aided Subset Sampling Based Motion Planner](https://mlsanity.com/paper/2410.00649/) (2024)

## Abstract (arXiv)

We compare the end-to-end training approach to a modular approach in which a system is decomposed into semantically meaningful components. We focus on the sample complexity aspect, in the regime where an extremely high accuracy is necessary, as is the case in autonomous driving applications. We demonstrate cases in which the number of training examples required by the end-to-end approach is exponentially larger than the number of examples required by the semantic abstraction approach.