AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Exploiting the Ground-Truth: An Adversarial Imitation Based Knowledge Distillation Approach for Event Detection

Jian Liu, Yubo Chen, Kang Liu

AAAI 2019 · Label-efficient and robust learning › Knowledge distillation · approach: Rethinking distillation objectives

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add event-argument embeddings to the teacher (ablation)Teacher + emb · ACE 2005 English goldenF1: 68.8 -> 74.2Improved
Add entity embeddings to the teacher (ablation)Teacher + emb · ACE 2005 English goldenF1: 68.8 -> 70.3Improved
Add entity embeddings on top of argument embeddings (ablation)Teacher + emb + agt · ACE 2005 English goldenF1: 74.2 -> 74.8Improved
Use adversarial imitation student instead of teacher on predicted tagsStudent-Final · ACE 2005 English predictedF1: 70.6 -> 71.2Improved
Use adversarial imitation student instead of ArgATTStudent-Final · ACE 2005 English predictedF1: 70.7 -> 71.2Improved
Use adversarial imitation student instead of CNNEDStudent-Final · ACE 2005 Chinese predictedF1: 57.8 -> 58.5Improved
Use predicted instead of golden annotations at test timeTeacher · ACE 2005 EnglishF1: 74.8 -> 70.6Hurt
Use predicted instead of golden annotations at test timeTeacher · ACE 2005 ChineseF1: 61.9 -> 57.3Hurt
Use heterogeneous attention teacher and CNN student (ablation)Student-Final · ACE 2005 English predictedF1: 71.2 -> 67.7Hurt
Use CNN for both teacher and student (ablation)Student-Final · ACE 2005 English predictedF1: 71.2 -> 70.8Hurt
Remove attention layer from student (ablation)Student-Final · ACE 2005 English predictedF1: 71.2 -> 68.6Hurt
Add the adversarial discriminator (vs none) (ablation)Student-Final · ACE 2005 English predictedF1: 69.2 -> 71.2Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy