Research map / Vision-language and multimodal / Robot action and navigation
Embodied-R1: Reinforced Embodied Reasoning for General Robotic Manipulation
Yifu Yuan, Haiqin Cui, Yaoting Huang, Yibin Chen, Fei Ni, Zibin Dong, Pengyi Li, YAN ZHENG, Hongyao Tang, Jianye HAO
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use RL (RFT) instead of SFT-only training (ablation) | Embodied-R1 3B · Where2Place | pointing accuracy: 41.25 -> 69.50 | Improved |
| use RL (RFT) instead of SFT-only training (ablation) | Embodied-R1 3B · Part-Afford | pointing accuracy: 40.20 -> 56.63 | Improved |
| use RL (RFT) instead of SFT-only training (ablation) | Embodied-R1 3B · VABench-V | MAE: 65.2 -> 45.0 | Improved |
| add ViRL common-sense data (ablation) | Embodied-R1 3B · spatial benchmarks | average rank: 3.4 -> 2.1 | Improved |
| add ViRL common-sense data (ablation) | Embodied-R1 3B · SAT Real | accuracy: 73.9 -> 76.3 | Improved |
| add RL after SFT on spatial reasoning (ablation) | Embodied-R1 3B · SAT Real | accuracy: 65.5 -> 76.3 | Improved |
| mix training data of multiple pointing abilities (ablation) | Embodied-R1 · Part-Afford | pointing accuracy: 51.25 -> 56.63 | Improved |
| mix training data of multiple pointing abilities (ablation) | Embodied-R1 · VABench-P | pointing accuracy: 65.39 -> 66.00 | Improved |
| add point number constraint reward (ablation) | Embodied-R1 · VABench-V | MAE: 59.7 -> 44.97 | Improved |
| use visual trace branch instead of affordance points | Embodied-R1 · real-world xArm | success rate (%): 83.3 -> 87.5 | Improved |
| integrate Embodied-R1 visual trace into Diffusion Policy | Diffusion Policy · LIBERO average | success rate (%): 64.4 -> 79.4 | Improved |
| use RGB-D dual-tower variant instead of RGB | Embodied-R1 · Open6DOR-Position overall | accuracy: 66.8 -> 94.3 | Improved |
Later papers that compared with this method
2 later accepted papers report Embodied-R1 as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models (NeurIPS 2026 · better on 2 of 2 shared results)
- E-TTS: A New Embodied Test-Time Scaling Framework for Robotic Manipulation (ECCV 2026 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Sofar — SoFar: Language-Grounded Orientation Bridges Spatial Reasoning and Object Manipulation (NeurIPS 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models (NeurIPS 2026 · similarity 0.92)
- Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics (NeurIPS 2025 · similarity 0.91)
- EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents (ICML 2025 · similarity 0.90)
- Robotic Visual Instruction (CVPR 2025 · similarity 0.89)
- Seeing Across Views: Benchmarking Spatial Reasoning of Vision-Language Models in Robotic Scenes (ICLR 2026 · similarity 0.89)
- ManiTaskGen: A Comprehensive Task Generator for Benchmarking and Improving Vision-Language Agents on Embodied Decision-Making (NeurIPS 2026 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
