Pith Number

pith:RE2WLRMI

pith:2026:RE2WLRMIB6W73PK2LHATYHJB7O

not attested not anchored not stored refs resolved

Hand-in-the-Loop: Improving Dexterous VLA via Seamless Interventional Correction

Liqun Huang, Nie Lin, Ruoshi Wen, Wei Xu, Xiao Ma, Xinjun Sheng, Zhengming Zhu, Zhuohang Li

HandITL blends human corrective intent with ongoing VLA policy execution to eliminate gesture jumps during dexterous hand takeovers.

arxiv:2605.15157 v1 · 2026-05-14 · cs.RO · 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{RE2WLRMIB6W73PK2LHATYHJB7O}

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

HandITL reduces takeover jitter by 99.8% and grasp failures by 87.5%, mean completion time by 19.1%, and produces policies that outperform those trained with standard teleoperation data by 19% on average across three long-horizon dexterous tasks.

C2weakest assumption

That the seamless blending of human corrective intent with autonomous policy execution can be achieved without introducing new instabilities or losing the benefit of the human correction in high-dimensional action spaces and contact-rich dynamics.

C3one line summary

HandITL blends human intent with policy execution to eliminate gesture jumps in dexterous VLA interventions, cutting jitter by 99.8%, grasp failures by 87.5%, and yielding 19% better refined policies.

References

34 extracted · 34 resolved · 6 Pith anchors

[1] Sample efficient interactive end-to-end deep learning for self-driving cars with selective multi-class safe dataset aggregation 2019

[2] GR00T N1: An Open Foundation Model for Generalist Humanoid Robots 2025 · arXiv:2503.14734

[3] $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control 2024 · arXiv:2410.24164

[4] GR-3 Technical Report 2025 · arXiv:2507.15493

[5] Conrft: A reinforced fine-tuning method for vla models via consistency policy 2025

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-17T21:40:25.426639Z
Last reissued	2026-05-17T21:57:18.749021Z
Builder	pith-number-builder-2026-05-17-v1
Signature	unsigned_v0
Schema	pith-number/v1.0

Canonical hash

893565c5880fadfdbd5a59c13c1d21fb908f22c45575bdb2b3b852d144fe4054

Aliases

arxiv: 2605.15157 · arxiv_version: 2605.15157v1 · pith_short_12: RE2WLRMIB6W7 · pith_short_16: RE2WLRMIB6W73PK2 · pith_short_8: RE2WLRMI

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/RE2WLRMIB6W73PK2LHATYHJB7O \
  | 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: 893565c5880fadfdbd5a59c13c1d21fb908f22c45575bdb2b3b852d144fe4054

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "8adf952441590f81f4c6b6d6af7f722a24f6bdc18c5e0b873c219eb7f4c2ec3e",
    "cross_cats_sorted": [
      "cs.LG"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.RO",
    "submitted_at": "2026-05-14T17:51:40Z",
    "title_canon_sha256": "c048f0a39f28d3aff1857ff1755136452a5c20b4481281a8372b6ab987fe90a5"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.15157",
    "kind": "arxiv",
    "version": 1
  }
}