Research map / Architectures and efficiency / Attention and transformers
Self-Attentive Hawkes Process
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 |
|---|---|---|---|
| Use SAHP rather than the strongest listed baseline | SAHP · Synthetic, RT, SOF, and MMC test sets | NLL per event: Synthetic: 1.43 -> 1.35; RT: 5.32 -> 4.56; SOF: 2.01 -> 1.86; MMC: 0.78 -> 0.52 | Helped |
| Use SAHP rather than the strongest listed baseline | SAHP · Synthetic, RT, SOF, and MMC test sets | F1 (%): Synthetic: 45.21 -> 58.50; RT: 45.25 -> 53.92; SOF: 6.34 -> 24.12; MMC: 34.00 -> 36.90 | Helped |
| Use SAHP rather than the strongest listed baseline | SAHP · Synthetic, RT, SOF, and MMC test sets | RMSE: Synthetic: 32.64 -> 31.16; RT: 1090.45 -> 1055.05; SOF: 154.13 -> 133.61; MMC: 4.12 -> 3.89 | Helped |
| Add time-shifted positional encoding instead of standard positional encoding (ablation) | SAHP · Synthetic, RT, SOF, and MMC test sets | F1 (%): Synthetic: 57.93 -> 58.50; RT: 53.24 -> 53.92; SOF: 24.05 -> 24.12; MMC: 34.23 -> 36.90 | Helped |
| Add time-shifted positional encoding instead of standard positional encoding (ablation) | SAHP · Synthetic, RT, SOF, and MMC test sets | RMSE: Synthetic: 33.32 -> 31.16; RT: 1102.34 -> 1055.05; SOF: 143.54 -> 133.61; MMC: 4.03 -> 3.89 | Helped |
| Increase Monte Carlo samples from 5 to 10 (ablation) | SAHP · SOF event prediction | Event prediction metrics | Helped |
| Increase Monte Carlo samples beyond 10 (ablation) | SAHP · SOF event prediction | Event prediction metrics | No clear effect |
Closest papers by meaning
- Hyper Hawkes Processes: Interpretable Models of Marked Temporal Point Processes (NeurIPS 2026 · similarity 0.86)
- EasyTPP: Towards Open Benchmarking Temporal Point Processes (ICLR 2024 · similarity 0.82)
- Understanding the Differences in Foundation Models: Attention, State Space Models, and Recurrent Neural Networks (NeurIPS 2024 · similarity 0.82)
- Memory Efficient Neural Processes via Constant Memory Attention Block (ICML 2024 · similarity 0.82)
- Training Dynamics of In-Context Learning in Linear Attention (ICML 2025 · similarity 0.82)
- Selective Induction Heads: How Transformers Select Causal Structures in Context (ICLR 2025 · similarity 0.82)
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.