Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Graph Mamba Operator: A Latent Simulator for Interacting Particle Systems

About

Modeling interacting dynamical systems requires capturing spatial interactions alongside long-range temporal dependencies. Graph neural networks (GNNs) provide a natural representation but typically rely on autoregressive rollouts and treat spatial and temporal dynamics separately, leading to error accumulation over long horizons. Existing approaches also focus on local interactions and short temporal contexts, limiting their ability to capture multi-hop dependencies and global structure. We introduce the Graph Mamba Operator (GraMO), a latent-space simulator that integrates state-space models with graph-based interaction learning. In contrast to prior work that sequences nodes or applies spatial and temporal updates in separate stages, GraMO couples graph-based interactions and temporal state updates within a single recurrence. The update is linear in the latent state, with input-dependent coefficients that adapt across regimes. We evaluate GraMO on N-body systems, motion capture, and robotics datasets, achieving the lowest error across benchmarks and the largest gains in long-horizon prediction.

Karn Tiwari, Niladri Dutta, N M Anoop Krishnan, Prathosh A P• 2026

Related benchmarks

TaskDatasetResultRank
State-to-Trajectory (S2T) PredictionMD17
S2T Error (Aspirin)0.51
15
Trajectory PredictionRope (test)
AMSE0.186
11
Trajectory PredictionSoft-robot (test)
AMSE (Scaled)0.036
11
Trajectory PredictionSwim Robotics (test)
AMSE0.11
11
Final-state predictionMD17
Aspirin Result0.74
10
Spatiotemporal Dynamics PredictionMoCap Run
FMSE (5 steps)0.023
10
Spatiotemporal Dynamics PredictionMoCap Walk
FMSE (5ts)0.039
9
Molecular Dynamics trajectory predictionAdK protein MD (test)
FMSE1.74
8
Showing 8 of 8 rows

Other info

Follow for update