[Adapt] [Seminar] Breaking NLI Systems with Sentences that Require Simple Lexical Inference

[hongru]

Hi Adapters, 
	
In this seminar, I’ll introduce the paper “Breaking NLI Systems with Sentences that Require Simple Lexical Inference (ACL 2018)”. 
This paper pointed out the limitation of the state-of-the-art Natural Langauge Inference (NLI) models by creating an adversarial datasets and designing thorough experiments, based on the essential analysis of the NLI problem itself. The method is simple and neat but not trivial and the results are convincing. 

Hope you can learn something from the talk =D
 
Time: Wed 4:30pm
Venue: SEIEE 3-414

Best regards,
Hongru

发送自 Windows 10 版邮件应用

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.sjtu.edu.cn/pipermail/adapt/attachments/20191105/80ec110c/attachment-0001.html>