Research map / Probabilistic and scientific modeling / Dynamical systems
STLnet: Signal Temporal Logic Enforced Multivariate Recurrent Neural Networks
Meiyi Ma, Ji Gao, Lu Feng, John Stankovic
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 |
|---|---|---|---|
| add STLnet training (student p) | LSTM · Synthetic phi5 | satisfaction rate: 0.84% -> 75.64% | Improved |
| add STLnet training (student p) | LSTM · Synthetic phi4 | satisfaction rate: 56.68% -> 83.09% | Improved |
| add STLnet training (student p) | Transformer · Synthetic phi1 | satisfaction rate: 27.76% -> 89.48% | Improved |
| add STLnet training (student p) | LSTM · Synthetic phi6 | satisfaction rate: 59.54% -> 61.85% | Improved |
| add STLnet training (student p) | LSTM · Synthetic phi5 | RMSE: 2.188 -> 1.151 | Improved |
| add STLnet training (student p) | Transformer · Synthetic phi2 | RMSE: 105.211 -> 111.688 | Hurt |
| add STLnet training (student p) | LSTM · Synthetic phi1 | RMSE: 0.026 -> 0.025 | Improved |
| use teacher STLnet-q projection at test time instead of student p | LSTM · Synthetic phi6 | RMSE: 8.532 -> 7.122 | Improved |
| use teacher STLnet-q projection at test time instead of student p | Transformer · Synthetic phi6 | RMSE: 8.326 -> 2.657 | Improved |
| use teacher STLnet-q projection at test time instead of student p | LSTM · Synthetic phi5 | satisfaction rate: 75.64% -> 100.00% | Improved |
| add STLnet | LSTM · Beijing air quality, prediction length 18 | RMSE: 162 -> 132 | Improved |
| add STLnet | LSTM · Beijing air quality | satisfaction rate: 20% -> over 70% | Improved |
Closest papers by meaning
- Resurrecting Recurrent Neural Networks for Long Sequences (ICML 2023 · similarity 0.85)
- Efficient Logit-based Knowledge Distillation of Deep Spiking Neural Networks for Full-Range Timestep Deployment (ICML 2025 · similarity 0.83)
- Neuro-Symbolic Temporal Point Processes (ICML 2024 · similarity 0.82)
- Exploiting Symmetric Temporally Sparse BPTT for Efficient RNN Training (AAAI 2024 · similarity 0.82)
- The Logical Expressiveness of Temporal GNNs via Two-Dimensional Product Logics (NeurIPS 2025 · similarity 0.82)
- TeLoGraF: Temporal Logic Planning via Graph-encoded Flow Matching (ICML 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.
