Pith Number

pith:H5VPMRH3

pith:2026:H5VPMRH3ZWEZIRVA3JURJ36NSU

not attested not anchored not stored refs resolved

ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both

Pheng-Ann Heng, Rain Liu, Xinyan Chen, Ziyu Guo

One discrete functional token suffices for both agentic operations and latent visual reasoning.

arxiv:2605.15198 v1 · 2026-05-14 · cs.CV · cs.AI · cs.CL

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{H5VPMRH3ZWEZIRVA3JURJ36NSU}

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

ATLAS achieves superior performance on challenging benchmarks while maintaining clear interpretability by using a single discrete functional token that serves as both an agentic operation and a latent visual reasoning unit.

C2weakest assumption

That a single discrete functional token can effectively internalize visual operations without any visual supervision and still generalize across tasks when generated via standard next-token prediction.

C3one line summary

ATLAS uses a single functional token to unify agentic and latent visual reasoning without image generation or external execution.

References

24 extracted · 24 resolved · 18 Pith anchors

[1] LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training · arXiv:2509.23661

[2] Qwen3-VL Technical Report · arXiv:2511.21631

[3] Anole: An open, autoregressive, native large multimodal models for interleaved image-text generation

[4] Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities · arXiv:2507.06261

[5] Emerging Properties in Unified Multimodal Pretraining · arXiv:2505.14683

Formal links

2 machine-checked theorem links

Cited by

1 paper in Pith

OpenCoF: Learning to Reason Through Video Generation

Receipt and verification

First computed	2026-05-17T21:40:24.984558Z
Last reissued	2026-05-17T21:57:18.377459Z
Builder	pith-number-builder-2026-05-17-v1
Signature	unsigned_v0
Schema	pith-number/v1.0

Canonical hash

3f6af644fbcd899446a0da6914efcd951095b5d33afb0a260046634726373acb

Aliases

arxiv: 2605.15198 · arxiv_version: 2605.15198v1 · pith_short_12: H5VPMRH3ZWEZ · pith_short_16: H5VPMRH3ZWEZIRVA · pith_short_8: H5VPMRH3

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/H5VPMRH3ZWEZIRVA3JURJ36NSU \
  | 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: 3f6af644fbcd899446a0da6914efcd951095b5d33afb0a260046634726373acb

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "c1c2ced4f400f5c1fce21efc76027fb18be7750b51ced415b4e1c73bacb6dbbc",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.CL"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-05-14T17:59:55Z",
    "title_canon_sha256": "402a185441e65509516ec4e2243c4b86cb4d33c0e299ea49cffc3afaf9d513b1"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.15198",
    "kind": "arxiv",
    "version": 1
  }
}