Research map / Probabilistic and scientific modeling / Variational and generative sampling
Fast and Flexible Temporal Point Processes with Triangular Maps
Oleksandr Shchur, Nicholas Gao, Marin Biloš, Stephan Günnemann
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace RNN with TriTPP for sampling long sequences | TPP model · Scalability, >400 events | sampling runtime: more than 100 times faster | Improved |
| replace RNN with TriTPP for training likelihood on short sequences | TPP model · Scalability, short sequences | training time per iteration | Hurt |
| replace RNN with TriTPP for training on sequences with more than 400 events | TPP model · Scalability | training time per iteration: more than 400 events | Improved |
| replace MRP with TriTPP (learnable block-diagonal transformations) | MRP · PUBG | test NLL: -0.83 -> -2.41 | Improved |
| replace MRP with TriTPP | MRP · SC | test NLL: 0.93 -> 0.83 | Improved |
| replace MRP with TriTPP | MRP · Hawkes1 | test NLL: 0.65 -> 0.56 | Improved |
| replace RP with TriTPP | RP · Hawkes2 | test NLL: 0.08 -> 0.00 | Improved |
| replace IPP with TriTPP | IPP · Twitter | test NLL: 1.60 -> 1.06 | Improved |
| replace RNN with TriTPP | RNN · PUBG | test NLL: -1.96 -> -2.41 | Improved |
| replace RNN with TriTPP | RNN · Reddit-S | test NLL: -4.89 -> -4.49 | Hurt |
| replace RNN with TriTPP | RNN · 13 datasets | test NLL: within 0.05 nats on 11 of 13 | No clear effect |
| replace RNN with TriTPP | RNN · IPP | MMD: 0.09 -> 0.02 | Improved |
Closest papers by meaning
- EasyTPP: Towards Open Benchmarking Temporal Point Processes (ICLR 2024 · similarity 0.91)
- Deep Continuous-Time State-Space Models for Marked Event Sequences (NeurIPS 2025 · similarity 0.89)
- Residual TPP: A Unified Lightweight Approach for Event Stream Data Analysis (ICML 2025 · similarity 0.88)
- Interacting Diffusion Processes for Event Sequence Forecasting (ICML 2024 · similarity 0.88)
- Decoupled Marked Temporal Point Process using Neural Ordinary Differential Equations (ICLR 2024 · similarity 0.87)
- Neural Jump-Diffusion Temporal Point Processes (ICML 2024 · 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.
