pith. machine review for the scientific record. sign in
Pith Number record

pith:LFILXV3B

pith:2018:LFILXV3BOO73RTABK7HNJRNJ54
Protocol content-addressed Pith receipt schema 1.0
Author
not author-attested
Timestamp
not externally anchored
Storage
storage not verified
References
references not resolved

QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation

Alexander Herzog, Alex Irpan, Deirdre Quillen, Dmitry Kalashnikov, Eric Jang, Ethan Holly, Julian Ibarz, Mrinal Kalakrishnan, Peter Pastor, Sergey Levine, Vincent Vanhoucke

Source: arxiv:1806.10293 v3 · Submitted 2018-06-27 · License http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Categories: cs.LG · cs.AI · cs.CV · cs.RO · stat.ML

Open Pith paper page Download JSON What is a Pith Number?

Record completeness checklist

Each row is a typed event attached to this Pith Number. Done rows show their third-party proof. Open rows have an inline form so you can complete them without leaving this page.

1 Anchor canonical hash to Bitcoin via OpenTimestamps open
Stamps the canonical SHA-256 (5950bbd76173bfb8cc0157ce...) across three calendars. Confirms instantly. Calendars upgrade to a Bitcoin-block proof in a few hours.
Free. No account. Anyone can re-stamp; each stamp is an independent calendar receipt.
2 Mirror to Internet Archive (Wayback Machine) open
Submits the paper URL to the Internet Archive and records the resulting Wayback capture URL plus timestamp. Pith does not host the paper.
Free, anonymous, third-party preservation.
3 Claim authorship open
Identity-backed. The same flow as the paper page. Sign in with ORCID, Apple, X, or email magic-link; if your ORCID is on the arXiv author list the attestation is one-click and shows ORCID verified.
Returns you right back here after sign-in.
4 Sign a citation pointing at this record open
Records a signed citation. Pick a relationship (supports, refutes, extends, uses_method, depends_on, replaces, compares, cites). Pith signs the assertion server-side.
More options
Held for moderation. Sign in to confirm immediately.
5 Submit replication, falsification, or formal bridge open
Public record of a replication attempt, formal bridge (Lean / Coq), or falsification. Held for moderation. Not peer review; a typed annotation.
More options
Held for moderation before it appears as confirmed.

Claims

Inbound citations

15 papers in Pith reference this work. 4 of those cite a specific Pith anchor.

GraspVLA: a Grasping Foundation Model Pre-trained on Billion-scale Synthetic Action Data
arXiv:2505.03233 · ref [55] · internal Pith anchor
MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations
arXiv:2310.17596 · ref [14] · internal Pith anchor
VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning
arXiv:2505.18719 · ref [35] · internal Pith anchor
HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Model
arXiv:2503.10631 · ref [84] · internal Pith anchor
World Action Models: The Next Frontier in Embodied AI
arXiv:2605.12090 · ref [118]
RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots
arXiv:2406.02523 · ref [18]
$\pi^{*}_{0.6}$: a VLA That Learns From Experience
arXiv:2511.14759 · ref [15]
LaST-R1: Reinforcing Robotic Manipulation via Adaptive Physical Latent Reasoning
arXiv:2604.28192 · ref [62]
CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
arXiv:2411.19650 · ref [27]
LaST-R1: Reinforcing Robotic Manipulation via Adaptive Physical Latent Reasoning
arXiv:2604.28192 · ref [62]
Open X-Embodiment: Robotic Learning Datasets and RT-X Models
arXiv:2310.08864 · ref [66]
Unmasking the Illusion of Embodied Reasoning in Vision-Language-Action Models
arXiv:2604.18000 · ref [60]
DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset
arXiv:2403.12945 · ref [26]
Octo: An Open-Source Generalist Robot Policy
arXiv:2405.12213 · ref [42]
OpenVLA: An Open-Source Vision-Language-Action Model
arXiv:2406.09246 · ref [46]

Receipt

First computed 2026-05-17T23:59:44.638710Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05, fingerprint 8d4b5ee74e4693bc…) · public key
Signature value oIvlzwcth5HQ5GpwD20ROTGywCPxywfN8GNNr9zHaUymwoys/eEWW6zKdjF2TLswrPZj/PO187VCu+jPDSRVAQ==
Schema pith-number/v1.0

Canonical hash

5950bbd76173bfb8cc0157ced4c5a9ef171730e95581f2cb61312f66c45a147d

Aliases

Kind Value
arxiv 1806.10293
arxiv_version 1806.10293v3
doi 10.48550/arxiv.1806.10293

Agent API

Same address serves humans and agents. Add an Accept header to get the JSON-LD payload at the resolver URL itself.

Resolver JSON Graph JSON Events JSON Schema Signing key Resolve alias

Verify this Pith Number yourself

Three shell lines recompute the canonical hash. If the hash you compute matches the one shown above, the record is intact end-to-end.

curl -sH 'Accept: application/ld+json' https://pith.science/pith/LFILXV3BOO73RTABK7HNJRNJ54 \
  | 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: 5950bbd76173bfb8cc0157ced4c5a9ef171730e95581f2cb61312f66c45a147d

Verify the Pith Ed25519 signature against the published public key:

curl -s https://pith.science/pith-signing-key.json | jq -r .public_key_b64 \
  | base64 -d > /tmp/pith_pub.bin
python3 -c "
from cryptography.hazmat.primitives.asymmetric.ed25519 import Ed25519PublicKey
import base64
pub = Ed25519PublicKey.from_public_bytes(open('/tmp/pith_pub.bin','rb').read())
sig = base64.b64decode('oIvlzwcth5HQ5GpwD20ROTGywCPxywfN8GNNr9zHaUymwoys/eEWW6zKdjF2TLswrPZj/PO187VCu+jPDSRVAQ==')
digest = bytes.fromhex('5950bbd76173bfb8cc0157ced4c5a9ef171730e95581f2cb61312f66c45a147d')
pub.verify(sig, digest)
print('signature OK')"
Show canonical record JSON 
{
  "metadata": {
    "abstract_canon_sha256": "bc71be13e337b79e3253f89afd9bc874d3edd4525dd4aa43752fe4558c2e29ff",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.CV",
      "cs.RO",
      "stat.ML"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2018-06-27T04:34:30Z",
    "title_canon_sha256": "d838e03fa1a7fd82ff1b77a79afb91a235d88797008c95a6565fbddf0a954fa0"
  },
  "schema_version": "1.0",
  "source": {
    "id": "1806.10293",
    "kind": "arxiv",
    "version": 3
  }
}