arbitration
modality-conflict-arbitration-v2
Modality-Conflict Arbitration Benchmark (v2)
A controlled benchmark for studying how a vision-language model arbitrates between
its two input channels when they disagree — and whether that choice tracks the
reliability of each channel.
Each row is a single conflict trial: an image of one math problem paired with the
text of a different problem. Because the two ground-truth answers are carried side by
side, the model's output alone tells you which modality it followed — no… See the full description on the dataset page: https://huggingface.co/datasets/vlm-modality-research/modality-conflict-arbitration-v2.han-decentralized-task-arbitration-dataset-v1
Humanoid Decentralized Task Arbitration Dataset
This dataset models task conflict resolution
and decentralized arbitration
between humanoid agents.
It captures competing task claims,
priority negotiation,
and consensus resolution outcomes.
Objective
To enable autonomous task arbitration
without centralized authority.
Data Fields
task_id
competing_agents
task_priority_score
arbitration_method
consensus_score
final_assigned_agent
resolution_latency
dispute_flag… See the full description on the dataset page: https://huggingface.co/datasets/achiepatricia/han-decentralized-task-arbitration-dataset-v1.arbitration_bad_example_2