Research map / Architectures and efficiency / RNNs and normalization
Learning Statistical Scripts with LSTM Recurrent Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace bigram event co-occurrence models with LSTM over event components | script learning · Narrative Cloze entities | R25: 0.124 -> 0.152 | Helped |
| replace bigram with LSTM (noun models) | script learning · Narrative Cloze nouns | R25: 0.037 -> 0.061 | Helped |
| add noun input to entity-predicting LSTM (ablation) | LSTM-ent-ent · Narrative Cloze entities | R25: 0.145 -> 0.152 | Helped |
| add entity ID input to noun-predicting LSTM (ablation) | LSTM-noun-noun · Narrative Cloze nouns | R25: 0.054 -> 0.061 | Helped |
| add entity rewriting to bigram model | All-Bigram · Narrative Cloze entities | R25: 0.124 -> 0.110 | Hurt |
| add subsequent-event conditioning (2D) to rewritten bigram | Rewrite Bigram · Narrative Cloze entities | R25: 0.110 -> 0.104 | Hurt |
| incorporate nouns into bigram model | All-Bigram · Crowdsourced inference (filtered) | Likert rating: 2.87 -> 2.21 | Hurt |
| predict nouns instead of entity IDs with LSTM | LSTM-both · Crowdsourced inference (filtered) | Likert rating: 3.08 -> 3.67 | Helped |
Closest papers by meaning
- Learning and Generalization in RNNs (NeurIPS 2021 · similarity 0.88)
- Learning Recurrent Binary/Ternary Weights (ICLR 2019 · similarity 0.87)
- Variable Computation in Recurrent Neural Networks (ICLR 2017 · similarity 0.87)
- Recurrent Batch Normalization (ICLR 2017 · similarity 0.87)
- Parallelizing Linear Recurrent Neural Nets Over Sequence Length (ICLR 2018 · similarity 0.87)
- Phased LSTM: Accelerating Recurrent Network Training for Long or Event-based Sequences (NeurIPS 2016 · similarity 0.87)
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.