# Chunking the Critic: A Transformer-based Soft Actor-Critic with N-Step Returns

arXiv: https://arxiv.org/abs/2503.03660 · HTML: https://mlsanity.com/paper/2503.03660/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- Placed in RL theory and sample complexity, one of 15 directions in Deep reinforcement learning methods.
- Builds on 51 of its 80 references that are in the map.
- First posted on arXiv 5 Mar 2025.

## Placed in

- Deep reinforcement learning methods / [RL theory and sample complexity](https://mlsanity.com/deeprl/rl-theory-and-sample-complexity/)

## Builds on (most cited first, 30 of 51)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016)
- [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor](https://mlsanity.com/paper/1801.01290/) (2018)
- [Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation](https://mlsanity.com/paper/1406.1078/) (2014)
- [Trust Region Policy Optimization](https://mlsanity.com/paper/1502.05477/) (2015)
- [Addressing Function Approximation Error in Actor-Critic Methods](https://mlsanity.com/paper/1802.09477/) (2018)
- [Asynchronous Methods for Deep Reinforcement Learning](https://mlsanity.com/paper/1602.01783/) (2016)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [Deep Reinforcement Learning with Double Q-learning](https://mlsanity.com/paper/1509.06461/) (2015)
- [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://mlsanity.com/paper/2106.01345/) (2021)
- [Offline Reinforcement Learning with Implicit Q-Learning](https://mlsanity.com/paper/2110.06169/) (2021)
- [Soft Actor-Critic Algorithms and Applications](https://mlsanity.com/paper/1812.05905/) (2018)
- [Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning](https://mlsanity.com/paper/1910.10897/) (2019)
- [IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures](https://mlsanity.com/paper/1802.01561/) (2018)
- [Deep Reinforcement Learning at the Edge of the Statistical Precipice](https://mlsanity.com/paper/2108.13264/) (2021)
- [Rainbow: Combining Improvements in Deep Reinforcement Learning](https://mlsanity.com/paper/1710.02298/) (2017)
- [Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning](https://mlsanity.com/paper/1910.00177/) (2019)
- [AWAC: Accelerating Online Reinforcement Learning with Offline Datasets](https://mlsanity.com/paper/2006.09359/) (2020)
- [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://mlsanity.com/paper/2106.02039/) (2021)
- [Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning](https://mlsanity.com/paper/1910.11956/) (2019)
- [Maximum a Posteriori Policy Optimisation](https://mlsanity.com/paper/1806.06920/) (2018)
- [IDQL: Implicit Q-Learning as an Actor-Critic Method with Diffusion Policies](https://mlsanity.com/paper/2304.10573/) (2023)
- [Stabilizing Transformers for Reinforcement Learning](https://mlsanity.com/paper/1910.06764/) (2019)
- [Safe and Efficient Off-Policy Reinforcement Learning](https://mlsanity.com/paper/1606.02647/) (2016)
- [Parameter Space Noise for Exploration](https://mlsanity.com/paper/1706.01905/) (2017)
- [TransDreamer: Reinforcement Learning with Transformer World Models](https://mlsanity.com/paper/2202.09481/) (2022)
- [Batch Renormalization: Towards Reducing Minibatch Dependence in Batch-Normalized Models](https://mlsanity.com/paper/1702.03275/) (2017)
- [Bigger, Regularized, Optimistic: scaling for compute and sample-efficient continuous control](https://mlsanity.com/paper/2405.16158/) (2024)

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

- [Reinforcement Learning with Action Chunking](https://mlsanity.com/paper/2507.07969/) (2025)
- [CO-RFT: Efficient Fine-Tuning of Vision-Language-Action Models through Chunked Offline Reinforcement Learning](https://mlsanity.com/paper/2508.02219/) (2025)
- [Actor-Critic for Continuous Action Chunks: A Reinforcement Learning Framework for Long-Horizon Robotic Manipulation with Sparse Reward](https://mlsanity.com/paper/2508.11143/) (2025)
- [DEAS: DEtached value learning with Action Sequence for Scalable Offline RL](https://mlsanity.com/paper/2510.07730/) (2025)
- [Multistep Quasimetric Learning for Scalable Goal-conditioned Reinforcement Learning](https://mlsanity.com/paper/2511.07730/) (2025)
- [SEAR: Sample Efficient Action Chunking Reinforcement Learning](https://mlsanity.com/paper/2603.01891/) (2026)
- [Adaptive Action Chunking via Multi-Chunk Q Value Estimation](https://mlsanity.com/paper/2605.10044/) (2026)
- [ReaCritic: Reasoning Transformer-based DRL Critic-model Scaling For Wireless Networks](https://mlsanity.com/paper/2505.10992/) (2025)
- [Adaptive Q-Chunking for Offline-to-Online Reinforcement Learning](https://mlsanity.com/paper/2605.05544/) (2026)
- [PAC-ACT: Post-training Actor-Critic for Action Chunking Transformers](https://mlsanity.com/paper/2607.09590/) (2026)
- [Sequence Aware SAC Control for Engine Fuel Consumption Optimization in Electrified Powertrain](https://mlsanity.com/paper/2508.04874/) (2025)
- [Whole-Body Mobile Manipulation using Offline Reinforcement Learning on Sub-optimal Controllers](https://mlsanity.com/paper/2604.12509/) (2026)
- [ACSAC: Adaptive Chunk Size Actor-Critic with Causal Transformer Q-Network](https://mlsanity.com/paper/2605.11009/) (2026)
- [Horizon Adaptive Offline Policy Learning via Value Stitching](https://mlsanity.com/paper/2606.21136/) (2026)
- [Upper-Expectile Multi-Step Q-Learning for Off-Policy Reinforcement Learning](https://mlsanity.com/paper/2608.02034/) (2026)
- [Action Chunking Proximal Policy Optimization with Feedback Correction](https://mlsanity.com/paper/2609.36250/) (2026)

## Abstract (arXiv)

We introduce a sequence-conditioned critic for Soft Actor-Critic (SAC) that models trajectory context with a lightweight Transformer and trains on aggregated N-step targets. Unlike prior approaches that (i) score state-action pairs in isolation or (ii) rely on actor-side action chunking to handle long horizons, our method strengthens the critic itself by conditioning on short trajectory segments and integrating multi-step returns -- without importance sampling (IS). The resulting sequence-aware value estimates capture the critical temporal structure for extended-horizon and sparse-reward problems. On local-motion benchmarks, we further show that freezing critic parameters for several steps makes our update compatible with CrossQ's core idea, enabling stable training without a target network. Despite its simplicity -- a 2-layer Transformer with 128-256 hidden units and a maximum update-to-data ratio (UTD) of 1 -- the approach consistently outperforms standard SAC and strong off-policy baselines, with particularly large gains on long-trajectory control. These results highlight the value of sequence modeling and N-step bootstrapping on the critic side for long-horizon reinforcement learning.