README.md

---
language:
- en
license: cc-by-nc-4.0
size_categories:
- 1K<n<10K
task_categories:
- image-text-to-text
- visual-question-answering
- multiple-choice
tags:
- vision-language
- multimodal
- benchmark
- chess
- chemistry
- music
- graph-theory
- semantic-equivalence
- VLM
dataset_info:
  features:
  - name: task
    dtype: string
  - name: domain
    dtype: string
  - name: index
    dtype: int32
  - name: question_type
    dtype: string
  - name: question
    dtype: string
  - name: notation
    dtype: string
  - name: notation_type
    dtype: string
  - name: option_a
    dtype: string
  - name: option_b
    dtype: string
  - name: option_c
    dtype: string
  - name: option_d
    dtype: string
  - name: correct_answer
    dtype: string
  - name: correct_idx
    dtype: int32
  - name: image
    dtype: image
  splits:
  - name: fork
    num_bytes: 0
    num_examples: 200
  - name: legal
    num_bytes: 0
    num_examples: 200
  - name: puzzle
    num_bytes: 0
    num_examples: 200
  - name: eval
    num_bytes: 0
    num_examples: 200
  - name: carbon
    num_bytes: 0
    num_examples: 200
  - name: hydrogen
    num_bytes: 0
    num_examples: 200
  - name: weight
    num_bytes: 0
    num_examples: 200
  - name: caption
    num_bytes: 0
    num_examples: 200
  - name: notes
    num_bytes: 0
    num_examples: 200
  - name: measures
    num_bytes: 0
    num_examples: 200
  - name: forms
    num_bytes: 0
    num_examples: 200
  - name: rhythm
    num_bytes: 0
    num_examples: 200
  - name: path_counting
    num_bytes: 0
    num_examples: 200
  - name: path_existence
    num_bytes: 0
    num_examples: 200
  - name: shortest_path
    num_bytes: 0
    num_examples: 200
  - name: bfs_traversal
    num_bytes: 0
    num_examples: 200
  download_size: 0
  dataset_size: 0
configs:
- config_name: default
  data_files:
  - split: fork
    path: data/fork-*
  - split: legal
    path: data/legal-*
  - split: puzzle
    path: data/puzzle-*
  - split: eval
    path: data/eval-*
  - split: carbon
    path: data/carbon-*
  - split: hydrogen
    path: data/hydrogen-*
  - split: weight
    path: data/weight-*
  - split: caption
    path: data/caption-*
  - split: notes
    path: data/notes-*
  - split: measures
    path: data/measures-*
  - split: forms
    path: data/forms-*
  - split: rhythm
    path: data/rhythm-*
  - split: path_counting
    path: data/path_counting-*
  - split: path_existence
    path: data/path_existence-*
  - split: shortest_path
    path: data/shortest_path-*
  - split: bfs_traversal
    path: data/bfs_traversal-*
---

# SEAM: Semantically Equivalent Across Modalities Benchmark for Vision-Language Models

*[CSSLab](https://csslab.cs.toronto.edu/), Department of Computer Science, University of Toronto*  
*[COLM '25] Second Conference on Language Modeling*

- **Paper**: [Paper](https://huggingface.co/papers/2508.18179)
- **Project Page / Leaderboard**: [SEAM Benchmark](https://lilv98.github.io/SEAM-Website/)
- **Code**: [GitHub](https://github.com/CSSLab/SEAM)

![Overview](https://lilv98.github.io/SEAM-Website/static/images/main.png)

## Abstract

Evaluating whether vision-language models (VLMs) reason consistently across representations is challenging because modality comparisons are typically confounded by task differences and asymmetric information. We introduce SEAM, a benchmark that pairs semantically equivalent inputs across four domains that have existing standardized textual and visual notations. By employing distinct notation systems across modalities, in contrast to OCR-based image-text pairing, SEAM provides a rigorous comparative assessment of the textual-symbolic and visual-spatial reasoning capabilities of VLMs. Across 21 contemporary models, we observe systematic modality imbalance: vision frequently lags language in overall performance, despite the problems containing semantically equivalent information, and cross-modal agreement is relatively low. Our error analysis reveals two main drivers: textual perception failures from tokenization in domain notation and visual perception failures that induce hallucinations. We also show that our results are largely robust to visual transformations. SEAM establishes a controlled, semantically equivalent setting for measuring and improving modality-agnostic reasoning.

## Key Features

- **4 Domains**: Chess, Chemistry, Music, Graph Theory with standardized notations
- **16 Tasks**: 4 tasks per domain (64 total task-modality combinations)
- **3 Modalities**: Language-only (L), Vision-only (V), Vision-Language (VL)
- **3,200 Base Samples**: 200 samples × 16 tasks
- **9,600 Evaluations**: TaskLoader generates 3 modality-specific prompts per base sample
- **Semantic Equivalence**: Same information presented in different representational formats

## Domains and Notation Systems

### Chess Domain
- **Tasks**: `fork`, `legal`, `puzzle`, `eval`
- **Textual**: FEN (Forsyth-Edwards Notation)
- **Visual**: Chess board diagrams

### Chemistry Domain  
- **Tasks**: `carbon`, `hydrogen`, `weight`, `caption`
- **Textual**: SMILES (Simplified Molecular Input Line Entry System)
- **Visual**: Chemical structure diagrams

### Music Domain
- **Tasks**: `notes`, `measures`, `forms`, `rhythm`
- **Textual**: ABC notation
- **Visual**: Musical staff notation

### Graph Theory Domain
- **Tasks**: `path_counting`, `path_existence`, `shortest_path`, `bfs_traversal`
- **Textual**: Adjacency matrices
- **Visual**: Node-edge diagrams

## Dataset Splits

The dataset is organized into 16 task-based splits (600 samples each):

- **Chess**: `fork`, `legal`, `puzzle`, `eval`
- **Chemistry**: `carbon`, `hydrogen`, `weight`, `caption`  
- **Music**: `notes`, `measures`, `forms`, `rhythm`
- **Graph Theory**: `path_counting`, `path_existence`, `shortest_path`, `bfs_traversal`

Each split contains 200 base samples. TaskLoader generates modality-specific prompts (L, V, VL) from these base samples.

## Usage

```python
from datasets import load_dataset

# Load the dataset
dataset = load_dataset("lilvjosephtang/SEAM-Benchmark")

# Access specific tasks
chess_fork = dataset["fork"]  # Chess fork detection (600 samples)
chemistry_carbon = dataset["carbon"]  # Carbon atom counting (600 samples)

# Each task contains 200 base samples
# TaskLoader generates modality-specific prompts (L/V/VL) from these base samples
print(f"Task {chess_fork[0]['task']} has {len(chess_fork)} base samples")

# Example sample structure
sample = chess_fork[0]
print(f"Task: {sample['task']}")
print(f"Domain: {sample['domain']}")
# No modality field - TaskLoader handles modality generation
print(f"Question: {sample['question']}")
print(f"Options: A) {sample['option_a']}, B) {sample['option_b']}, C) {sample['option_c']}, D) {sample['option_d']}")
print(f"Correct Answer: {sample['correct_answer']}")
print(f"Notation: {sample['notation']}")  # FEN string for chess
# sample['image'] contains the chess board image for Vision/Vision-Language modalities
```

## Sample Structure

Each sample contains:
- `task`: Task identifier (e.g., "fork", "carbon")
- `domain`: Domain category ("chess", "chemistry", "music", "graph")
- No modality field (TaskLoader generates modality-specific prompts)
- `index`: Sample index within the task
- `question`: Question text (if applicable)
- `notation`: Domain-specific notation (FEN, SMILES, ABC, adjacency matrix)
- `notation_type`: Type of notation used
- `option_a`, `option_b`, `option_c`, `option_d`: Multiple choice options
- `correct_answer`: The correct answer
- `correct_idx`: Index of the correct option
- `image`: Associated image (PIL Image, None for base storage - TaskLoader handles image loading for V/VL modalities)

## Evaluation Protocol

SEAM enables three types of evaluation:

1.  **Language**: Models receive only textual notation
2.  **Vision**: Models receive only visual representation  
3.  **Vision-Language**: Models receive both notation and image

The semantic equivalence across modalities allows for direct comparison of reasoning capabilities and cross-modal agreement analysis.

## Citation

```bibtex
@inproceedings{
  tang2025seam,
  title={{SEAM}: Semantically Equivalent Across Modalities Benchmark for Vision-Language Models},
  author={Zhenwei Tang and Difan Jiao and Blair Yang and Ashton Anderson},
  booktitle={Second Conference on Language Modeling},
  year={2025},
  url={https://openreview.net/forum?id=lI4LgGv4sX}
}
@misc{tang2025seamsemanticallyequivalentmodalities,
      title={SEAM: Semantically Equivalent Across Modalities Benchmark for Vision-Language Models}, 
      author={Zhenwei Tang and Difan Jiao and Blair Yang and Ashton Anderson},
      year={2025},
      eprint={2508.18179},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2508.18179}, 
}
```