Upload folder using huggingface_hub

README.md

@@ -0,0 +1,84 @@
+---
+dataset_info:
+  features:
+    - name: ID
+      dtype: int64
+    - name: Game
+      dtype: string
+    - name: tier
+      dtype: int64
+    - name: Question
+      dtype: string
+    - name: Answer
+      dtype: string
+    - name: game_state_url
+      dtype: string
+configs:
+  - config_name: default
+    data_files:
+      - split: test
+        path: ludobench.parquet
+license: mit
+task_categories:
+  - visual-question-answering
+  - question-answering
+language:
+  - en
+size_categories:
+  - n<1K
+pretty_name: LudoBench
+tags:
+  - board-games
+  - multimodal-reasoning
+  - benchmark
+---
+
+# LudoBench
+
+**LLMs as Rules Oracles: Exploring Real-World Multimodal Reasoning in Tabletop Strategy Game Environments**
+
+*ICLR 2026*
+
+[![Paper](https://img.shields.io/badge/Paper-OpenReview-blue)](https://openreview.net/forum?id=TOgQ00DEek)
+[![Dataset](https://img.shields.io/badge/Dataset-HuggingFace-yellow)](https://huggingface.co/datasets/launch/LudoBench)
+[![GitHub](https://img.shields.io/badge/Code-GitHub-black)](https://github.com/jpeper/LudoBench)
+[![Demo](https://img.shields.io/badge/Demo-HF%20Space-orange)](https://huggingface.co/spaces/launch/LudoBench)
+
+---
+
+A multimodal board-game reasoning benchmark evaluating LLM/VLM reasoning
+across 5 strategy games and 3 difficulty tiers.
+
+## Dataset Description
+
+- **638** annotated question-answer pairs
+- **5 games**: Kingdomino, Res Arcana, Pax Renaissance, Carcassonne, Catan
+- **3 tiers**: Environment Perception (T1), Rules Integration (T2), Short-Horizon Optimization (T3)
+- **3 rules modalities**: None (parametric), Text (text rulebook), Image (image rulebook)
+
+## Fields
+
+| Field | Description |
+|-------|-------------|
+| `ID` | Unique question identifier |
+| `Game` | Board game name |
+| `tier` | Difficulty tier (1, 2, or 3) |
+| `Question` | The question text |
+| `Answer` | Expected answer |
+| `game_state_url` | Path(s) to game state image(s), semicolon-separated if multiple |
+
+## Benchmark Results
+
+See `benchmark_results.csv` for accuracy scores of 9 models across all game/tier/modality splits.
+
+## Citation
+
+```bibtex
+@inproceedings{peper2026ludobench,
+  title={{LLMs} as Rules Oracles: Exploring Real-World Multimodal Reasoning in Tabletop Strategy Game Environments},
+  author={Peper, Joseph J. and Gandra, Sai Krishna and Zhang, Yunxiang and Chennareddy, Vaibhav and Jha, Shloki and Payani, Ali and Wang, Lu},
+  booktitle={Proceedings of the Fourteenth International Conference on Learning Representations (ICLR)},
+  year={2026},
+  address={Rio de Janeiro, Brazil}
+}
+```

benchmark_results.csv

@@ -0,0 +1,28 @@
+row,KingD | None,KingD | Text,KingD | Image,Res Arcana | None,Res Arcana | Text,Res Arcana | Image,Pax Ren. | None,Pax Ren. | Text,Pax Ren. | Image,Carca. | None,Carca. | Text,Carca. | Image,Catan | None,Catan | Text,Catan | Image
+GPT-4o | T1,0.381,0.571,0.429,0.65,0.65,0.575,0.375,0.475,0.6,0.425,0.35,0.4,0.45,0.45,0.45
+o1 | T1,0.524,0.429,0.429,0.675,0.525,0.6,0.45,0.55,0.475,0.45,0.45,0.375,0.475,0.55,0.55
+GPT-4.1 | T1,0.619,0.524,0.619,0.775,0.725,0.75,0.525,0.575,0.6,0.575,0.4,0.475,0.575,0.45,0.65
+o3 | T1,0.75,0.675,0.65,0.775,0.775,0.7,0.475,0.675,0.65,0.45,0.5,0.575,0.6,0.575,0.55
+GPT-5.1 | T1,0.75,0.65,0.675,0.775,0.8,0.8,0.6,0.6,0.65,0.725,0.525,0.55,0.6,0.575,0.575
+Gemini 2.5 Flash | T1,0.524,0.476,0.381,0.675,0.775,0.5,0.55,0.4,0.375,0.45,0.4,0.35,0.625,0.5,0.325
+Gemini 2.5 Pro | T1,0.524,0.524,0.333,0.65,0.7,0.525,0.65,0.725,0.375,0.425,0.35,0.45,0.575,0.65,0.425
+Gemini 3 Pro | T1,0.825,0.8,0.85,0.875,0.85,0.775,0.775,0.75,0.8,0.825,0.75,0.65,0.525,0.625,0.7
+Claude 4.5 Sonnet | T1,0.571,0.619,0.524,0.65,0.8,0.8,0.6,0.65,,0.375,0.325,0.3,0.6,0.625,
+GPT-4o | T2,0.3,0.433,0.3,0.225,0.4,0.35,0.125,0.525,0.475,0.125,0.15,0.225,0.225,0.2,0.2
+o1 | T2,0.2,0.4,0.267,0.35,0.35,0.35,0.25,0.4,0.45,0.225,0.25,0.3,0.15,0.25,0.225
+GPT-4.1 | T2,0.333,0.433,0.3,0.4,0.5,0.475,0.275,0.4,0.375,0.175,0.2,0.15,0.325,0.225,0.325
+o3 | T2,0.35,0.4,0.375,0.325,0.625,0.475,0.3,0.55,0.575,0.275,0.275,0.275,0.375,0.275,0.275
+GPT-5.1 | T2,0.3,0.325,0.275,0.325,0.525,0.525,0.2,0.6,0.467,0.25,0.25,0.325,0.275,0.275,0.3
+Gemini 2.5 Flash | T2,0.167,0.3,0.267,0.225,0.3,0.3,0.25,0.3,0.375,0.25,0.3,0.2,0.25,0.275,0.125
+Gemini 2.5 Pro | T2,0.367,0.367,0.267,0.375,0.475,0.375,0.1,0.4,0.25,0.225,0.275,0.15,0.275,0.35,0.175
+Gemini 3 Pro | T2,0.725,0.675,0.75,0.55,0.625,0.65,0.325,0.475,0.425,0.425,0.425,0.325,0.5,0.325,0.5
+Claude 4.5 Sonnet | T2,0.3,0.333,0.233,0.433,0.633,0.4,0.3,0.533,,0.275,0.275,0.375,0.225,0.275,
+GPT-4o | T3,0.0,0.0,0.0,0.02,0.06,0.0,0.053,0.053,0.018,0.04,0.0,0.04,0.065,0.0,0.032
+o1 | T3,0.0,0.02,0.0,0.04,0.06,0.04,0.105,0.018,0.053,0.08,0.04,0.04,0.032,0.065,0.032
+GPT-4.1 | T3,0.04,0.08,0.1,0.02,0.02,0.02,0.123,0.123,0.088,0.02,0.0,0.06,0.194,0.065,0.032
+o3 | T3,0.06,0.12,0.04,0.0,0.06,0.06,0.07,0.175,0.211,0.12,0.06,0.1,0.194,0.226,0.258
+GPT-5.1 | T3,0.16,0.06,0.06,0.08,0.08,0.08,0.123,0.088,0.211,0.02,0.1,0.1,0.129,0.161,0.161
+Gemini 2.5 Flash | T3,0.0,0.02,0.0,0.0,0.08,0.08,0.018,0.053,0.0,0.02,0.06,0.04,0.032,0.0,0.065
+Gemini 2.5 Pro | T3,0.04,0.02,0.0,0.12,0.3,0.08,0.123,0.053,0.088,0.06,0.06,0.04,0.129,0.097,0.161
+Gemini 3 Pro | T3,0.16,0.16,0.14,0.2,0.26,0.18,0.053,0.105,0.193,0.06,0.02,0.04,0.129,0.161,0.032
+Claude 4.5 Sonnet | T3,0.02,0.04,0.04,0.04,0.06,0.08,0.053,0.088,,0.04,0.02,0.06,0.0,0.032,

ludobench.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57085d48a31e89b0fa050b7dd923d90ccb84dd5c1bfd94c3f357a0f731f14a2f
+size 43073