paper-with-me

홈 › Papers

SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learning

2024-10-13 · Hojoon Lee, Dongyoon Hwang, Donghu Kim, Hyunseung Kim, Jun Jet Tai, Kaushik Subramanian, Peter R. Wurman, Jaegul Choo, Peter Stone, Takuma Seno

Recent advances in CV and NLP have been largely driven by scaling up the number of network parameters, despite traditional theories suggesting that larger networks are prone to overfitting. These large networks avoid overfitting by integrating components that induce a simplicity bias, guiding models toward simple and generalizable solutions. However, in deep RL, designing and scaling up networks have been less explored. Motivated by this opportunity, we present SimBa, an architecture designed to scale up parameters in deep RL by injecting a simplicity bias. SimBa consists of three components: (i) an observation normalization layer that standardizes inputs with running statistics, (ii) a residual feedforward block to provide a linear pathway from the input to output, and (iii) a layer normalization to control feature magnitudes. By scaling up parameters with SimBa, the sample efficiency of various deep RL algorithms-including off-policy, on-policy, and unsupervised methods-is consistently improved. Moreover, solely by integrating SimBa architecture into SAC, it matches or surpasses state-of-the-art deep RL methods with high computational efficiency across DMC, MyoSuite, and HumanoidBench. These results demonstrate SimBa's broad applicability and effectiveness across diverse RL algorithms and environments.

📄 PDF Abstract BibTeX arXiv:2410.09754

Code (4)

sonyresearch/simba 공식 구현 jax
MindSpore-scientific/code-8/tree/main/Simba mindspore
araffin/sbx jax
pwc-1/Paper-9/tree/main/7/Simba mindspore

Tasks

Computational EfficiencyDeep Reinforcement Learning

Methods 이 논문이 사용한 방법론

Dilated Convolution 설명 없음
Average Pooling 설명 없음
1x1 Convolution A 1 x 1 Convolution is a convolution with some special properties in that it can be used for dimensionality reduction,…
Convolution A convolution is a type of matrix operation, consisting of a kernel, a small matrix of weights, that slides over input data performing element-wise multiplication with the…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Global Average Pooling Global Average Pooling is a pooling operation designed to replace fully connected layers in classical CNNs. The idea is to generate one feature map for each corresponding…
SAC 설명 없음

Similar Papers 제목 키워드 기반

Hyperspherical Normalization for Scalable Deep Reinforcement Learning

2025-02-21 · Hojoon Lee, Youngdo Lee, Takuma Seno, Donghu Kim 외

Scaling up the model size and computation has brought consistent performance improvements in supervised learning. However, this lesson often fails to apply to reinforcement learning (RL) because training the model on non…

continuous-controlContinuous ControlDeep Reinforcement Learningreinforcement-learning+2

A Forget-and-Grow Strategy for Deep Reinforcement Learning Scaling in Continuous Control

2025-07-03 · Zilin Kang, Chenyuan Hu, Yu Luo, Zhecheng Yuan 외 arxiv

Deep reinforcement learning for continuous control has recently achieved impressive progress. However, existing methods often suffer from primacy bias, a tendency to overfit early experiences stored in the replay buffer,…

Reinforcement LearningContinuous Control

SSIMBaD: Sigma Scaling with SSIM-Guided Balanced Diffusion for AnimeFace Colorization

2025-06-04 · Junpyo Seo, Hanbin Koo, Jieun Yook, Byung-Ro Moon

We propose a novel diffusion-based framework for automatic colorization of Anime-style facial sketches. Our method preserves the structural fidelity of the input sketch while effectively transferring stylistic attributes…

ColorizationSSIM

R2R2: Robust Representation for Intensive Experience Reuse via Redundancy Reduction in Self-Predictive Learning

2026-05-13 · Sanghyeob Song, Donghyeok Lee, Jinsik Kim, Sungroh Yoon arxiv

For reinforcement learning in data-scarce domains like real-world robotics, intensive data reuse enhances efficiency but induces overfitting. While prior works focus on critic bias, representation-level instability in Se…

Reinforcement LearningContinuous Control

SiMBA: Simplified Mamba-Based Architecture for Vision and Multivariate Time series

2024-03-22 · Badri N. Patro, Vijay S. Agneeswaran

Transformers have widely adopted attention networks for sequence mixing and MLPs for channel mixing, playing a pivotal role in achieving breakthroughs across domains. However, recent literature highlights issues with att…

Inductive BiasMambaState Space ModelsTime Series+1