Pith Number

pith:FYN4RFFB

pith:2024:FYN4RFFBVFSWUUUGMZ3R5ZSEP2

not attested not anchored not stored refs resolved

RT-H: Action Hierarchies Using Language

Debidatta Dwibedi, Dorsa Sadigh, Jonathan Tompson, Pierre Sermanet, Quon Vuong, Suneel Belkhale, Ted Xiao, Tianli Ding, Yevgen Chebotar

Predicting fine-grained language descriptions of motions first helps robot policies share structure across diverse tasks and accept language corrections.

arxiv:2403.01823 v2 · 2024-03-04 · cs.RO · cs.AI

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

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

Our method RT-H builds an action hierarchy using language motions: it first learns to predict language motions, and conditioned on this and the high-level task, it predicts actions, using visual context at all stages.

C2weakest assumption

That fine-grained language motion phrases capture shared low-level structure across semantically diverse tasks sufficiently well that predicting them improves downstream action prediction and enables effective language-based correction.

C3one line summary

RT-H learns robot policies by first predicting language motions as an intermediate representation and then mapping those plus the high-level task to actions, yielding more robust multi-task performance and the ability to learn from language interventions.

References

63 extracted · 63 resolved · 6 Pith anchors

[1] Do as i can, not as i say: Grounding language in robotic affordances 2023

[2] “No, to the Right 2023 · doi:10.1145/3568162.3578623

[3] Correcting robot plans with natural language feedback

[4] URL https://api.semanticscholar.org/CorpusID: 248085271

[5] RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control 2023 · arXiv:2307.15818

Formal links

2 machine-checked theorem links

Cited by

21 papers in Pith

Reflection-Based Task Adaptation for Self-Improving VLA

A Survey on Vision-Language-Action Models: An Action Tokenization Perspective

AVA-VLA: Improving Vision-Language-Action models with Active Visual Attention

Continually Evolving Skill Knowledge in Vision Language Action Model

DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge

Receipt and verification

First computed	2026-05-17T23:38:14.778993Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

2e1bc894a1a9656a528666771ee6447eb469f837ee05bea1f488a7363f760f38

Aliases

arxiv: 2403.01823 · arxiv_version: 2403.01823v2 · doi: 10.48550/arxiv.2403.01823 · pith_short_12: FYN4RFFBVFSW · pith_short_16: FYN4RFFBVFSWUUUG · pith_short_8: FYN4RFFB

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/FYN4RFFBVFSWUUUGMZ3R5ZSEP2 \
  | 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: 2e1bc894a1a9656a528666771ee6447eb469f837ee05bea1f488a7363f760f38

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "92b344e6172175696b4c494f461c8d89070242449cfbc5c528c263a6933d5a7a",
    "cross_cats_sorted": [
      "cs.AI"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.RO",
    "submitted_at": "2024-03-04T08:16:11Z",
    "title_canon_sha256": "6ce889699add44e9d8826eff1f6ba9e286a1dd915a3843ad177b00c773c46637"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2403.01823",
    "kind": "arxiv",
    "version": 2
  }
}