Pith Number

pith:IEKNEQTO

pith:2024:IEKNEQTOVQGQLCXB6M3UXO3Q6P

not attested not anchored not stored refs resolved

MathVerse: Does Your Multi-modal LLM Truly See the Diagrams in Visual Math Problems?

Aojun Zhou, Dongzhi Jiang, Haokun Lin, Hongsheng Li, Kai-Wei Chang, Pan Lu, Peng Gao, Pengshuo Qiu, Renrui Zhang, Yichi Zhang, Ziyu Guo

MathVerse shows multi-modal LLMs often solve visual math problems using text rather than diagrams.

arxiv:2403.14624 v2 · 2024-03-21 · cs.CV · 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{IEKNEQTOVQGQLCXB6M3UXO3Q6P}

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

Current visual math benchmarks contain excessive textual information that allows MLLMs to deduce answers without truly interpreting the input diagrams; MathVerse's multi-version design enables equitable evaluation of whether and how much MLLMs understand visual diagrams for mathematical reasoning.

C2weakest assumption

The human-annotated transformations into six versions with varying degrees of multi-modal information accurately isolate visual understanding without introducing new biases or inconsistencies in problem difficulty or meaning.

C3one line summary

MathVerse is a benchmark that tests multi-modal LLMs on visual math by providing each problem in six versions with progressively less diagram and text information to measure true visual understanding.

References

83 extracted · 83 resolved · 26 Pith anchors

[1] Advances in Neural Information Processing Systems 35, 23716–23736 (2022) 2022

[2] arXiv preprint arXiv:1905.13319 , year= 1905 · arXiv:1905.13319

[3] Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond 2023 · arXiv:2308.12966

[4] In: Advances in neural information processing systems 1901

[5] In: Proceedings of the 29th International Conference on Computa- tional Linguistics 2022

Formal links

2 machine-checked theorem links

Cited by

26 papers in Pith

ErrorRadar: Benchmarking Complex Mathematical Reasoning of Multimodal Large Language Models Via Error Detection

MathFlow: Enhancing the Perceptual Flow of MLLMs for Visual Mathematical Problems

Draw2Think: Harnessing Geometry Reasoning through Constraint Engine Interaction

mPLUG-Owl3: Towards Long Image-Sequence Understanding in Multi-Modal Large Language Models

Beyond Mode Collapse: Distribution Matching for Diverse Reasoning

Receipt and verification

First computed	2026-05-17T23:38:45.982351Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

4114d2426eac0d058ae1f3374bbb70f3d118e7729ca5530f450ef123757c0850

Aliases

arxiv: 2403.14624 · arxiv_version: 2403.14624v2 · doi: 10.48550/arxiv.2403.14624 · pith_short_12: IEKNEQTOVQGQ · pith_short_16: IEKNEQTOVQGQLCXB · pith_short_8: IEKNEQTO

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/IEKNEQTOVQGQLCXB6M3UXO3Q6P \
  | 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: 4114d2426eac0d058ae1f3374bbb70f3d118e7729ca5530f450ef123757c0850

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "bd2d944b86d298df85475113d4d792ced4626b8f40ed7ecd677b10b501932bf1",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.CL",
      "cs.LG"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2024-03-21T17:59:50Z",
    "title_canon_sha256": "87352c938c6afc2dea3ac47841997565364989bda833e8bd9032db8ca5aa4b05"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2403.14624",
    "kind": "arxiv",
    "version": 2
  }
}