Pith Number

pith:ABUITWGJ

pith:2025:ABUITWGJPFUECEEDS6XA226RDY

not attested not anchored not stored refs resolved

SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning

Bowen Zhou, Dehui Wang, Dingxiang Luo, Ganqu Cui, Haozhan Li, Jiale Yu, Jiangmiao Pang, Jia Zeng, Kaiyan Zhang, Ning Ding, Shanghang Zhang, Tianxing Chen, Xuekai Zhu, Yao Mu, Youbang Sun, Yuchen Fan, Yuchen Zhang, Yuhao Zhang, Yu Wang, Yuxin Zuo, Zhaohui Yang

Reinforcement learning scales vision-language-action model training beyond supervised fine-tuning

arxiv:2509.09674 v1 · 2025-09-11 · cs.RO · cs.AI · cs.CL · cs.LG

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{ABUITWGJPFUECEEDS6XA226RDY}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

When applied to OpenVLA-OFT, SimpleVLA-RL achieves SoTA performance on LIBERO and even outperforms π₀ on RoboTwin 1.0&2.0 with the exploration-enhancing strategies we introduce. SimpleVLA-RL not only reduces dependence on large-scale data and enables robust generalization, but also remarkably surpasses SFT in real-world tasks.

C2weakest assumption

That the introduced VLA-specific trajectory sampling, multi-environment rendering, and exploration-enhancing strategies remain stable and effective across different base VLA models and real-world distribution shifts without extensive additional tuning or hidden failure modes.

C3one line summary

SimpleVLA-RL applies tailored reinforcement learning to VLA models, reaching SoTA on LIBERO, outperforming π₀ on RoboTwin, and surpassing SFT in real-world tasks while reducing data needs and identifying a 'pushcut' phenomenon.

References

44 extracted · 44 resolved · 30 Pith anchors

[1] OpenVLA: An Open-Source Vision-Language-Action Model · arXiv:2406.09246

[2] A Survey on Vision-Language-Action Models: An Action Tokenization Perspective · arXiv:2507.01925

[3] Roumelio- tis, and Manoj Karkee

[4] What Matters in Learning from Offline Human Demonstrations for Robot Manipulation · arXiv:2108.03298

[5] AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems · arXiv:2503.06669

Formal links

2 machine-checked theorem links

Cited by

38 papers in Pith

VGAS: Value-Guided Action-Chunk Selection for Few-Shot Vision-Language-Action Adaptation

Agentic-VLA: Efficient Online Adaptation for Vision-Language-Action Models

Towards Long-Lived Robots: Continual Learning VLA Models via Reinforcement Fine-Tuning

TwinRL: Digital Twin-Driven Reinforcement Learning for Real-World Robotic Manipulation

RankQ: Offline-to-Online Reinforcement Learning via Self-Supervised Action Ranking

Receipt and verification

First computed	2026-05-17T23:38:53.074646Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

006889d8c9796841108397ae0d6bd11e09207c2fd66e155720ba682bbbdc6b27

Aliases

arxiv: 2509.09674 · arxiv_version: 2509.09674v1 · doi: 10.48550/arxiv.2509.09674 · pith_short_12: ABUITWGJPFUE · pith_short_16: ABUITWGJPFUECEED · pith_short_8: ABUITWGJ

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/ABUITWGJPFUECEEDS6XA226RDY \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 006889d8c9796841108397ae0d6bd11e09207c2fd66e155720ba682bbbdc6b27

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "eb6febe5fd595ea05910a580fc224467f6772dcd2bd2f7323867fe71e413ff2e",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.CL",
      "cs.LG"
    ],
    "license": "http://creativecommons.org/licenses/by-nc-sa/4.0/",
    "primary_cat": "cs.RO",
    "submitted_at": "2025-09-11T17:59:17Z",
    "title_canon_sha256": "705bc96580cb34c4cd93cf4d2b6f80daaae6ea990a9552a70416de12a645416b"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2509.09674",
    "kind": "arxiv",
    "version": 1
  }
}