Research map / Architectures and efficiency / Attention and transformers
Transformer Embeddings of Irregularly Spaced Events and Their Participants
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 NHP's LSTM with continuous-time attention | NHP · MIMIC-II | held-out log-likelihood, next-event time RMSE, and event-type error rate | Mixed |
| Replace NHP's LSTM with continuous-time attention | NHP · StackOverflow | held-out log-likelihood, next-event time RMSE, and event-type error rate | Helped |
| Replace THP, SAHP, or NHP with A-NHP | A-NHP · Synthetic | next-event timing and type prediction | Mixed |
| Replace NHP with A-NHP | NHP · RoboCup | held-out log-likelihood and event-type prediction | Helped |
| Add the RoboCup NDTT domain rules | NHP and A-NHP · RoboCup | held-out log-likelihood and event-type error rate | Helped |
| Replace NDTT's LSTM with A-NDTT attention | NDTT · RoboCup | overall accuracy | No clear effect |
| Add domain rules, attention, or both | NHP · IPTV | held-out log-likelihood | Helped |
| Add domain rules | NHP · IPTV | event-type error rate | Helped |
| Add LayerNorm and a two-layer feed-forward network (ablation) | continuous-time Transformer · Pilot experiments | model performance | No clear effect |
| Use the simplified attention architecture instead of the traditional Transformer (ablation) | continuous-time Transformer · Pilot experiments | model performance | No clear effect |
Closest papers by meaning
- Hyper Hawkes Processes: Interpretable Models of Marked Temporal Point Processes (NeurIPS 2026 · similarity 0.86)
- Selective Induction Heads: How Transformers Select Causal Structures in Context (ICLR 2025 · similarity 0.83)
- Neuro-Symbolic Temporal Point Processes (ICML 2024 · similarity 0.83)
- Rough Transformers: Lightweight and Continuous Time Series Modelling through Signature Patching (NeurIPS 2024 · similarity 0.82)
- Interacting Diffusion Processes for Event Sequence Forecasting (ICML 2024 · similarity 0.82)
- Gridded Transformer Neural Processes for Spatio-Temporal Data (ICML 2025 · similarity 0.81)
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.