AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Neural Question Generation with Answer Pivot

Bingning Wang, Xiaochuan Wang, Ting Tao, Qi Zhang, Jingfang Xu

AAAI 2020 · Large language models › Retrieval and knowledge · approach: Explanatory and multiple-choice QA benchmarks · 29 citations (OpenAlex)

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 answer pivot layer (hidden answer predictor) (ablation)Transformer question generator, answer-agnostic · SQuAD NQG testBLEU-4: 13.09 -> 14.78Improved
add answer pivot layer (ablation)Transformer question generator · SQuAD NQG testMeteor: 17.24 -> 18.61Improved
add answer pivot layer (ablation)Transformer question generator · SQuAD NQG testRouge-L: 40.98 -> 41.87Improved
add answer pivot layer (ablation)Transformer question generator · Human evaluationAverage rating: 3.63 -> 3.97Improved
add supervised answer pivot loss L_S with ground-truth answers (ablation)answer pivot model · Human evaluationAverage rating: 3.97 -> 4.14Improved
add supervised answer pivot loss L_S (ablation)answer pivot model · Answer pivot qualityF1: 66.2 -> 83.9Improved
add question generation objective (joint model) vs answer-only sequence labeling (ablation)answer pivot predictor · Answer pivot qualityF1: 80.5 -> 83.9Improved
add auxiliary pivot prediction loss L_APanswer pivot model · SQuAD NQGposterior collapse of pivotImproved
answer pivot model vs answer-aware baselines (pivot predicted rather than given)NQG · SQuAD NQG answer-awareBLEU-4: 15.27 -> 16.42Improved
answer pivot model vs NQG++ on answerabilityNQG · Human evaluationAnswerable: 4.64 -> 3.98Hurt
use generated questions from our model as additional MRC training data vs NQG++ and SynNet dataBERT base on SQuAD · SQuAD MRCEM / F1Improved

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