Add WAM paper: SIMPLE: Simulation-Based Policy Learning and Evaluation for Humanoid Loco-manipulation

[279object]

Paper

• Title: SIMPLE: Simulation-Based Policy Learning and Evaluation for Humanoid Loco-manipulation
• Short name: SIMPLE
• arXiv ID: 2606.08278v1
• Paper: https://arxiv.org/pdf/2606.08278
• Authors: Songlin Wei, Zhenhao Ni, Jie Liu, Zhenyu Zhao, Junjie Ye, Hongyi Jing, Junkai Xia, Xiawei Liu, Michael Leong, Liang Heng, Di Huang, Yue Wang
• Published: 2026-06-06
• arXiv categories: cs.RO
• Matched keywords: world action model, world action models

README Entry

- **SIMPLE**: "SIMPLE: Simulation-Based Policy Learning and Evaluation for Humanoid Loco-manipulation", arXiv 2026.
  [[📄 Paper](https://arxiv.org/pdf/2606.08278)]

Robot Decision

• Decision: accept
• Confidence: medium
• Taxonomy: evaluation_action_policy_bimanual_humanoid
• Taxonomy confidence: high
• README heading: #### 🤖 Action Policy — Bimanual and Humanoid Form
• Badges: None

Reason

The paper is primarily a humanoid simulation benchmark, but it explicitly integrates and evaluates recent World Action Models for embodied humanoid loco-manipulation.

Evidence

• benchmarks recent world action models (WAMs)
• native compatibility with WAMs like DreamZero
• policy evaluation for humanoid loco-manipulation

Taxonomy Evidence

• unified simulation testbed for humanoid policy learning and evaluation
• 60 diverse whole-body tasks across 50 indoor scenes
• benchmarks VLA models and recent WAMs
• automated trajectory generation and VR teleoperation pipelines

Human Review Checklist

• This paper belongs in the WAM survey
• README section is correct
• Badges are correct
• Short name is correct
• Paper link is correct
• Merge this PR if accepted