AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

TopicEq: A Joint Topic and Mathematical Equation Model for Scientific Texts

Michihiro Yasunaga, John D. Lafferty

AAAI 2019 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Disentangled and discrete latent models

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
jointly train the context topic model with an equation LSTMTopicEq topic model · ContextEq-400K test, 50 topicstopic coherence (NPMI): .085 -> .097; +0.01Improved
jointly train the context topic model with an equation LSTMTopicEq topic model · ContextEq-400K test, 100 topicstopic coherence (NPMI): .084 -> .094Improved
add equations as bags of tokens instead of omitting equation input (ablation)TopicEq topic model · ContextEq-400K test, 50 topicstopic coherence (NPMI): .085 -> .087; +0.002Improved
replace equation LSTM with equation bag-of-tokens model (ablation)TopicEq topic model · ContextEq-400K test, 50 topicstopic coherence (NPMI): .097 -> .087Hurt
shuffle equation token order (ablation)TopicEq topic model · ContextEq-400K test, 50 topicstopic coherence (NPMI): .097 -> .086Hurt
condition equation LSTM on topic informationEquation generation · ContextEq-400K test, 100 topicsequation perplexity: 5.81 -> 5.34Improved
condition equation LSTM on topic informationEquation generation · ContextEq-400K test, 100 topicssyntax error rate: 15.3 -> 11.7Improved
embed the topic vector inside the LSTM cell rather than adding it at the outputTopic-dependent equation LSTM · ContextEq-400K test, 100 topicsequation perplexity: 5.41 -> 5.34Improved
embed the topic vector inside the LSTM cell rather than adding it at the outputTopic-dependent equation LSTM · ContextEq-400K test, 100 topicssyntax error rate: 12.5 -> 11.7Improved
make symbol-to-word alignment depend on topicMath-symbol phrase prediction · ContextEq-400K test, 50 topicsphrase prediction perplexity: 602 -> 406Improved
make symbol-to-word alignment depend on topicMath-symbol phrase prediction · ContextEq-400K test, 100 topicsphrase prediction perplexity: 602 -> 387Improved
add joint topic-aware alignment learningTopicEq topic model · ContextEq-400K test, 50 topicstopic coherence (NPMI): .085 -> .088Improved

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