Zebra Logic Bench - a allenai Collection

allenai 's Collections

MolmoAct2 Eval Rollouts

Molmo2-ER Datasets

MolmoAct2-BimanualYAM Dataset

MolmoAct2 Datasets

MolmoAct2 Finetuned Models

MolmoAct2 Models

Branch-Adapt-Route

intent-aware-lfqa

MolmoPoint-Data

MolmoBot-Models

Open Coding Agents Specialization

Open Coding Agents

Olmo 3 Pre-training

Olmo 3 Post-training

MolmoAct Data Mixture

OLMoE (January 2025)

Tulu 3 Datasets

OLMoE (November 2024)

Tulu V2.5 Suite

AI2 Safety Toolkit

Zebra Logic Bench

OLMo 2 Preview Post-trained Models

Zebra Logic Bench

updated Dec 23, 2025

ZebraLogic Bench: Testing the Limits of LLMs in Logical Reasoning

Running

Agents

95

Zebra Logic Bench

🦓

95

Display model leaderboard and explore sample puzzles
allenai/ZebraLogicBench

Viewer • Updated Jul 11, 2024 • 4.26k • 1.67k • 26

Note 1k puzzles in grid format and 3k in mcqa format. answers are hidden.
allenai/ZebraLogicBench-private

Viewer • Updated Jul 4, 2024 • 4.26k • 521 • 14

Note answers are included here. to prevent data leakage, plz apply for access.
Faith and Fate: Limits of Transformers on Compositionality

Paper • 2305.18654 • Published May 29, 2023 • 9

Note Our NeurIPS 2023 (spotlight) paper about the data and analysis.