DailyDialog++

Multi-reference Adversarial Dataset for Dialog Evaluation

Explore DailyDialog++

Version 1.0, test set

Visualize Model Predictions

RUBER-Large (random negatives) (Peking University)

RUBER-Large (random + adversarial negatives) (Peking University)

RUBER (random negatives) (Peking University)

RUBER (random + adversarial negatives) (Peking University)

DEB (random negatives) (IIT Madras)

DEB (random + adversarial negatives) (IIT Madras)