Research map / Reinforcement learning / Policy gradient and actor-critic
Deep Reinforcement Learning of Marked Temporal Point Processes
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 TPPRL to schedule reviews by optimizing sampled recall at test time | Spaced repetition · Duolingo-based simulated student model | recall at test time | Helped |
| Use TPPRL to choose post times without assuming reverse-chronological feed ordering | Smart broadcasting · Algorithmically sorted feeds | average rank | Helped |
| Use TPPRL to choose post times without assuming reverse-chronological feed ordering | Smart broadcasting · Algorithmically sorted feeds | time at the top | Helped |
| Force the policy to be piecewise constant between events by setting w_t = 0 (ablation) | TPPRL · Smart broadcasting, algorithmically sorted feeds | relative decrease in average rank: 33% (mean) and 33% (median) -> 28% (mean) and 30% (median) | Hurt |
| Force the policy to be piecewise constant between events by setting w_t = 0 (ablation) | TPPRL · Smart broadcasting, algorithmically sorted feeds | increase in time at the top: about 11% -> 9% | Hurt |
| Force the policy to be piecewise constant between events by setting w_t = 0 (ablation) | TPPRL · Spaced repetition | mean recall at test time: 38.9% -> 37.9% | Hurt |
| Use TPPRL instead of RedQueen | Smart broadcasting · Reverse-chronological feeds | average rank | No clear effect |
| Use TPPRL instead of Karimi's method | Smart broadcasting · Reverse-chronological feeds | time at the top | Helped |
Closest papers by meaning
- Learning Temporal Point Processes via Reinforcement Learning (NeurIPS 2018 · similarity 0.90)
- Decoupled Marked Temporal Point Process using Neural Ordinary Differential Equations (ICLR 2024 · similarity 0.85)
- Not Only Where, But When: Temporal Scheduling for RLVR (NeurIPS 2026 · similarity 0.85)
- Deep Continuous-Time State-Space Models for Marked Event Sequences (NeurIPS 2025 · similarity 0.84)
- Dense Reward for Free in Reinforcement Learning from Human Feedback (ICML 2024 · similarity 0.84)
- TTRL: Test-Time Reinforcement Learning (NeurIPS 2025 · similarity 0.84)
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.