DinanE. x 9
bookRéférences 9
The second conversational intelligence challenge (convai2)
We describe the setting and results of the ConvAI2 NeurIPS competition that aims to further the stat...
2019-11-30 00:00:00
Build it break it fix it for dialogue safety: Robustness from adversarial human attack
… 2017), employing hate speech that expresses prejudice against a particular group, or &a...
2019-08-17 00:00:00
Facebook AI Research {edinan, roller, kshuster, angelafan, michaelauli, jase}@ fb. com
In open-domain dialogue intelligent agents should exhibit the use of knowledge, however there are fe...
SafetyKit: First aid for measuring safety in open-domain conversational systems
… 2021), we use sentiment analysis with VADER (… 2014), a rule-based sentiment...
Anticipating safety issues in e2e conversational ai: Framework and tooling
… They measure fairness as outcome discrepancies (such as politeness or sentiment) with w...
Reducing conversational agents' overconfidence through linguistic calibration
While improving neural dialogue agents’ factual accuracy is the object of much research, a...
… What a fine place this is.’, ’I am so happy for this day.Even if is...
… We demonstrate the technique using a set of finegrained sentiment labels such as happy,...
Controlling style in generated dialogue
… We do not make use of the sentiment labels from ED in this work. They might be somewhat...