# TQL: Scaling Q-Functions with Transformers by Preventing Attention Collapse

arXiv: https://arxiv.org/abs/2602.01439 · HTML: https://mlsanity.com/paper/2602.01439/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Deep RL algorithms and representations, one of 15 directions in Deep reinforcement learning methods.
- Builds on 35 of its 40 references that are in the map.
- First posted on arXiv 1 Feb 2026.

## Placed in

- Deep reinforcement learning methods / [Deep RL algorithms and representations](https://mlsanity.com/deeprl/deep-rl-algorithms-and-representations/)

## Builds on (most cited first, 30 of 35)

- [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor](https://mlsanity.com/paper/1801.01290/) (2018)
- [Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems](https://mlsanity.com/paper/2005.01643/) (2020)
- [Conservative Q-Learning for Offline Reinforcement Learning](https://mlsanity.com/paper/2006.04779/) (2020)
- [GLU Variants Improve Transformer](https://mlsanity.com/paper/2002.05202/) (2020)
- [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://mlsanity.com/paper/2106.01345/) (2021)
- [Root Mean Square Layer Normalization](https://mlsanity.com/paper/1910.07467/) (2019)
- [Offline Reinforcement Learning with Implicit Q-Learning](https://mlsanity.com/paper/2110.06169/) (2021)
- [Off-Policy Deep Reinforcement Learning without Exploration](https://mlsanity.com/paper/1812.02900/) (2018)
- [A Minimalist Approach to Offline Reinforcement Learning](https://mlsanity.com/paper/2106.06860/) (2021)
- [CogView: Mastering Text-to-Image Generation via Transformers](https://mlsanity.com/paper/2105.13290/) (2021)
- [Scaling Vision Transformers to 22 Billion Parameters](https://mlsanity.com/paper/2302.05442/) (2023)
- [AWAC: Accelerating Online Reinforcement Learning with Offline Datasets](https://mlsanity.com/paper/2006.09359/) (2020)
- [Behavior Regularized Offline Reinforcement Learning](https://mlsanity.com/paper/1911.11361/) (2019)
- [Offline Reinforcement Learning as One Big Sequence Modeling Problem](https://mlsanity.com/paper/2106.02039/) (2021)
- [MOReL : Model-Based Offline Reinforcement Learning](https://mlsanity.com/paper/2005.05951/) (2020)
- [IDQL: Implicit Q-Learning as an Actor-Critic Method with Diffusion Policies](https://mlsanity.com/paper/2304.10573/) (2023)
- [OGBench: Benchmarking Offline Goal-Conditioned RL](https://mlsanity.com/paper/2410.20092/) (2024)
- [Flow Q-Learning](https://mlsanity.com/paper/2502.02538/) (2025)
- [Small-scale proxies for large-scale Transformer training instabilities](https://mlsanity.com/paper/2309.14322/) (2023)
- [Spectral Norm Regularization for Improving the Generalizability of Deep Learning](https://mlsanity.com/paper/1705.10941/) (2017)
- [Bigger, Better, Faster: Human-level Atari with human-level efficiency](https://mlsanity.com/paper/2305.19452/) (2023)
- [Stop Regressing: Training Value Functions via Classification for Scalable Deep RL](https://mlsanity.com/paper/2403.03950/) (2024)
- [Q-Transformer: Scalable Offline Reinforcement Learning via Autoregressive Q-Functions](https://mlsanity.com/paper/2309.10150/) (2023)
- [Uncertainty Weighted Actor-Critic for Offline Reinforcement Learning](https://mlsanity.com/paper/2105.08140/) (2021)
- [Revisiting the Minimalist Approach to Offline Reinforcement Learning](https://mlsanity.com/paper/2305.09836/) (2023)
- [Bigger, Regularized, Optimistic: scaling for compute and sample-efficient continuous control](https://mlsanity.com/paper/2405.16158/) (2024)
- [Generalized Decision Transformer for Offline Hindsight Information Matching](https://mlsanity.com/paper/2111.10364/) (2021)
- [Mixtures of Experts Unlock Parameter Scaling for Deep RL](https://mlsanity.com/paper/2402.08609/) (2024)
- [1000 Layer Networks for Self-Supervised RL: Scaling Depth Can Enable New Goal-Reaching Capabilities](https://mlsanity.com/paper/2503.14858/) (2025)
- [Stable Gradients for Stable Learning at Scale in Deep Reinforcement Learning](https://mlsanity.com/paper/2506.15544/) (2025)

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

- [EXPO-FT: Sample-Efficient Reinforcement Learning Finetuning for Vision-Language-Action Models](https://mlsanity.com/paper/2605.25477/) (2026)
- [Reinforcement Learning via Value Gradient Flow](https://mlsanity.com/paper/2604.14265/) (2026)
- [FASTER: Value-Guided Sampling for Fast RL](https://mlsanity.com/paper/2604.19730/) (2026)
- [What Does Flow Matching Bring To TD Learning?](https://mlsanity.com/paper/2603.04333/) (2026)
- [Do You Really Need to Pretrain Q-Functions for Online RL Fine-Tuning?](https://mlsanity.com/paper/2607.27203/) (2026)
- [Q-Learning With World Models](https://mlsanity.com/paper/2608.17163/) (2026)
- [Reinforcement Learning for Real-Time Vision-Language-Action Policies](https://mlsanity.com/paper/2609.18207/) (2026)
- [ACSAC: Adaptive Chunk Size Actor-Critic with Causal Transformer Q-Network](https://mlsanity.com/paper/2605.11009/) (2026)
- [Looped Actor: Depth-Recurrent Reasoning Models for Reinforcement Learning](https://mlsanity.com/paper/2609.37432/) (2026)

## Abstract (arXiv)

Despite scale driving substantial recent advancements in machine learning, reinforcement learning (RL) methods still primarily use small value functions. Naively scaling value functions -- including with a transformer architecture, which is known to be highly scalable -- often results in learning instability and worse performance. In this work, we ask what prevents transformers from scaling effectively for value functions? Through empirical analysis, we identify the critical failure mode in this scaling: attention scores collapse as capacity increases. Our key insight is that we can effectively prevent this collapse and stabilize training by controlling the entropy of the attention scores, thereby enabling the use of larger models. To this end, we propose Transformer Q-Learning (TQL), a method that unlocks the scaling potential of transformers in learning value functions in RL. Our approach yields up to a 43% improvement in performance when scaling from the smallest to the largest network sizes, while prior methods suffer from performance degradation.