Meta-Reinforcement Learning with Discrete World Models for Adaptive Load Balancing

ACMSE 25 2025 43.8 method, application

TLDR

Integrates meta-RL with DreamerV3's discrete world models for adaptive load balancing, outperforming A2C and resisting catastrophic forgetting.

Reasoning

The paper presents a novel combination of meta-reinforcement learning and world models for load balancing, demonstrating strong adaptation and resilience. However, the abstract lacks detailed experimental setup and comparisons to other world model approaches, limiting assessment of generalizability.

Read-first score

Read-first score 43.8, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 32.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Topical relevance 42%

45.7

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=none

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 397.

Keyword Scores

world model

model-based reinforcement learning world model

world dynamics prediction

generative world model

world simulator

interactive world model

video world model

Deep Analysis

Innovations

Integration of meta-reinforcement learning with DreamerV3 architecture for load balancing in operating systems
Rapid adaptation to dynamic workloads with minimal retraining
Robust resilience to catastrophic forgetting under varying workload distributions and sizes

Methodology

The approach combines a meta-reinforcement learning algorithm with the DreamerV3 architecture, trained on load balancing tasks in operating systems. It is evaluated against the Advantage Actor-Critic (A2C) algorithm in both standard and adaptive trials, measuring performance under varying workload distributions and sizes.

Key Results

The proposed method outperforms A2C in standard and adaptive trials, maintaining high performance and demonstrating robust resilience to catastrophic forgetting across dynamic workload conditions.

Resources

Paper