AI Paper Atlas

Research map / Probabilistic and scientific modeling / Dynamical systems

Neural Hybrid Automata: Learning Dynamics With Multiple Modes and Stochastic Transitions

Michael Poli, Stefano Massaroli, Luca Scimeca, Sanghyuk Chun, Seong Joon Oh, Atsushi Yamashita, Hajime Asama, Jinkyoo Park, Animesh Garg

NeurIPS 2021 · Probabilistic and scientific modeling › Dynamical systems · approach: Switching and law-discovery dynamics · 7 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use NHA mode encoder instead of DBSCANNHA · TCP mode recovery, m=3v-measure: 0.66 -> 0.86Improved
Use NHA mode encoder instead of k-means++NHA · TCP mode recovery, m=10v-measure: 0.30 -> 0.96Improved
Increase number of latent modes m from 3 to 10 (ablation)NHA · TCP mode recoveryv-measure: 0.86 -> 0.96Improved
Increase number of supervising trajectories n from 1 to 10 (ablation)NHA event module · TCP event moduleflow NLL: 2.761 -> 2.313Improved
Increase number of supervising trajectories n from 1 to 10 (ablation)NHA event module · TCP event modulejump map MSE (1e-3): 1.435 -> 0.003Improved
Increase n from 3 to 10 (ablation)NHA event module · TCP event moduleflow NLL: 2.375 -> 2.313No clear effect
Add segmentation noise p=0.5 (ablation)NHA-5 · TCP mode recoveryv-measure: 0.91 -> 0.43Hurt
Add segmentation noise p=0.3 (ablation)NHA-3 · TCP mode recoveryv-measure: 0.86 -> 0.73Hurt
Add segmentation noise p=0.5 (ablation)DBSCAN · TCP mode recoveryv-measure: 0.69 -> 0.23Hurt
Add segmentation noise p=0.1 (ablation)NHA-3 · TCP mode recoveryv-measure: +0.02No clear effect
Use softmax latents instead of categorical sampling (ablation)NHA encoder · Switching linear systemmode recovery qualityHurt
Use one extra latent mode beyond system modes (overclustering) (ablation)NHA · Training stabilitytraining stabilityImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy