Research map / Probabilistic and scientific modeling / Molecules and proteins
Neural Dynamic Focused Topic Model
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 |
|---|---|---|---|
| decouple topic activity from topic proportion using time-dependent Bernoulli topic masks | NDF-TM · UN, NeurIPS, ACL | PPL-DC: UN: 1 748 $7$ 13 -> 1 527 $7$ 36; NeurIPS: 7 746$7$699 -> 6 529$7$26; ACL: 7 805$7$182 -> 7 690$7$215 | Helped |
| decouple topic activity from topic proportion using time-dependent Bernoulli topic masks | NDF-TM · UN, NeurIPS, ACL | P-NLL: UN: 7.615 $7$ 0.005 -> 7.640$7$0.004; NeurIPS: 8.983 $7$ 0.003 -> 8.901 $7$ 0.001; ACL: 8.84 $7$ 0.02 -> 8.88 $7$ | Mixed |
| use neural rather than identity transition functions for temporal latent variables (ablation) | NDF-TM · UN, NeurIPS, ACL | PPL-DC: UN: 1 578 $7$ 29 -> 1 527 $7$ 36; NeurIPS: 6 549$7$21 -> 6 529$7$26; ACL: 7 877$7$213 -> 7 690$7$215 | Helped |
| use neural rather than identity transition functions for temporal latent variables (ablation) | NDF-TM · UN, NeurIPS, ACL | P-NLL: UN: 7.682$7$0.080 -> 7.640$7$0.004; NeurIPS: 8.923 $7$ 0.002 -> 8.901 $7$ 0.001; ACL: 8.91 $7$ 0.03 -> 8.88 $7$ 0. | Mixed |
| use neural rather than identity transition functions for temporal latent variables (ablation) | NDF-TM · UN, NeurIPS, ACL | TC: UN: 0.43 $7$ 0.18 -> 0.46 $7$ 0.20; NeurIPS: -0.53$7$0.02 -> -0.50$7$0.04; ACL: -0.74$7$0.11 -> -0.64$7$0.12 | Mixed |
| use neural rather than identity transition functions for temporal latent variables (ablation) | NDF-TM · UN, NeurIPS, ACL | TD: UN: 0.56 $7$ 0.03 -> 0.63$7$ 0.01; NeurIPS: 0.90$7$0.01 -> 0.85$7$0.02; ACL: 0.73$7$0.01 -> 0.74$7$0.01 | Mixed |
Closest papers by meaning
- A Discrete Variational Recurrent Topic Model without the Reparametrization Trick (NeurIPS 2020 · similarity 0.91)
- Discovering Discrete Latent Topics with Neural Variational Inference (ICML 2017 · similarity 0.90)
- Document Informed Neural Autoregressive Topic Models with Distributional Prior (AAAI 2019 · similarity 0.88)
- Neural Topic Model via Optimal Transport (ICLR 2021 · similarity 0.88)
- Dynamic Topic Models for Temporal Document Networks (ICML 2022 · similarity 0.88)
- textTOvec: DEEP CONTEXTUALIZED NEURAL AUTOREGRESSIVE TOPIC MODELS OF LANGUAGE WITH DISTRIBUTED COMPOSITIONAL PRIOR (ICLR 2019 · similarity 0.88)
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.