# True Online Temporal-Difference Learning

arXiv: https://arxiv.org/abs/1512.04087 · HTML: https://mlsanity.com/paper/1512.04087/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Dec 2015.

## Built on by (most cited first, 16 of 16)

- [Streaming Deep Reinforcement Learning Finally Works](https://mlsanity.com/paper/2410.14606/) (2024)
- [An Analysis of Frame-skipping in Reinforcement Learning](https://mlsanity.com/paper/2102.03718/) (2021)
- [Finite-Sample Analysis of Off-Policy TD-Learning via Generalized Bellman Operators](https://mlsanity.com/paper/2106.12729/) (2021)
- [Deep Reinforcement Learning-based Intelligent Traffic Signal Controls with Optimized CO2 emissions](https://mlsanity.com/paper/2310.13129/) (2023)
- [An Auction-based Marketplace for Model Trading in Federated Learning](https://mlsanity.com/paper/2402.01802/) (2024)
- [Reinforcement Learning for Flow-Matching Policies with Density Transport](https://mlsanity.com/paper/2606.08602/) (2026)
- [Sim-Env: Decoupling OpenAI Gym Environments from Simulation Models](https://mlsanity.com/paper/2102.09824/) (2021)
- [DeepADMR: A Deep Learning based Anomaly Detection for MANET Routing](https://mlsanity.com/paper/2302.13877/) (2023)
- [Two-Step Q-Learning](https://mlsanity.com/paper/2407.02369/) (2024)
- [Revisiting Adam for Streaming Reinforcement Learning](https://mlsanity.com/paper/2605.06764/) (2026)
- [BP(λ): Online Learning via Synthetic Gradients](https://mlsanity.com/paper/2401.07044/) (2024)
- [A Multi-Step Minimax Q-learning Algorithm for Two-Player Zero-Sum Markov Games](https://mlsanity.com/paper/2407.04240/) (2024)
- [Hierarchical Reinforcement Learning Framework for Adaptive Walking Control Using General Value Functions of Lower-Limb Sensor Signals](https://mlsanity.com/paper/2507.16983/) (2025)
- [Swift-Sarsa: Fast and Robust Linear Control](https://mlsanity.com/paper/2507.19539/) (2025)
- [Intentional Updates for Streaming Reinforcement Learning](https://mlsanity.com/paper/2604.19033/) (2026)
- [rl-triton: High-Performance Triton GPU Kernels for Reinforcement Learning Credit Assignment](https://mlsanity.com/paper/2608.17641/) (2026)

## Abstract (arXiv)

The temporal-difference methods TD(λ) and Sarsa(λ) form a core part of modern reinforcement learning. Their appeal comes from their good performance, low computational cost, and their simple interpretation, given by their forward view. Recently, new versions of these methods were introduced, called true online TD(λ) and true online Sarsa(λ), respectively (van Seijen & Sutton, 2014). These new versions maintain an exact equivalence with the forward view at all times, whereas the traditional versions only approximate it for small step-sizes. We hypothesize that these true online methods not only have better theoretical properties, but also dominate the regular methods empirically. In this article, we put this hypothesis to the test by performing an extensive empirical comparison. Specifically, we compare the performance of true online TD(λ)/Sarsa(λ) with regular TD(λ)/Sarsa(λ) on random MRPs, a real-world myoelectric prosthetic arm, and a domain from the Arcade Learning Environment. We use linear function approximation with tabular, binary, and non-binary features. Our results suggest that the true online methods indeed dominate the regular methods. Across all domains/representations the learning speed of the true online methods are often better, but never worse than that of the regular methods. An additional advantage is that no choice between traces has to be made for the true online methods. Besides the empirical results, we provide an in-depth analysis of the theory behind true online temporal-difference learning. In addition, we show that new true online temporal-difference methods can be derived by making changes to the online forward view and then rewriting the update equations.