Latent Dynamics · 2026-10-09
AtomWorld-Mirror: Macro-Step World Modeling of Critical Evolution Backbones for Materials Dynamics
Ziming Pan, Ruge Zhang, Haozhi Han, Junkai Zhou, Xingyuan Chen, Yifeng Chen, Yunquan Zhang, Ting Cao, Yunxin Liu, Kun Li
AI-assisted summary based on the paper. Consult the linked source for methods and limitations. Benchmark numbers appear after editor verification.
TL;DR
AtomWorld-Mirror is a time-aware macro-step world model for long-horizon atomistic materials evolution. It distills short micro-event segments into physically reachable transitions between key “critical evolution backbone” states, jointly predicting sparse structural edits and accumulated continuous-time. The model enforces local reachability, inventory conservation, and continuous-time consistency, and replaces explicit micro-event replay with macro-step inference for large speedups.
Why it matters
Long-horizon atomistic simulation is limited by an “evolutionary-resolution bottleneck”: most computation resolves low-impact local updates before structurally decisive states occur. AtomWorld-Mirror targets that bottleneck by predicting backbone-level sparse state-time transitions directly, while preserving structural validity and time semantics through explicit physical constraints tied to the underlying simulator’s reachable dynamics.
Method
- Distill step-wise atomistic teachers into time-aware macro-step transitions over a critical evolution backbone, predicting sparse reachable edits plus accumulated physical time.
- Constrain each macro transition with local reachability (edits restricted to teacher-derived candidate supports), inventory conservation (projection preserves material counts), and continuous-time consistency (duration is path-conditioned).
- At rollout, evaluate candidate macro horizons, discard illegal candidates, and select a horizon to advance both structure and clock without explicit micro-event replay.
Limitation
AtomWorld-Mirror targets long-horizon materials evolution in regimes where explicit microscopic replay consumes most of the computational budget before structurally consequential states are reached; the current operating range reaches k = 1024.
Abstract (from arXiv)
Atomistic simulation is a fundamental tool for studying long-term materials evolution, from diffusion and defect dynamics to interfacial reactions and fracture. Yet conventional simulators typically advance at microscopic resolution, spending substantial computation on low-impact local updates before reaching structurally consequential states, an evolutionary-resolution bottleneck that limits long-horizon simulation. We propose AtomWorld-Mirror, a time-aware macro-step world model for the critical evolution backbone of atomic systems. For Step-Wise atomistic simulation, AtomWorld-Mirror distills short micro-event segments into physically reachable transitions between key states, jointly predicting sparse structural edits and accumulated physical time through latent macro-step dynamics. Local reachability, inventory conservation, and continuous-time consistency constrain each transition. By amortizing local atomic physics into a reusable latent macro model and replacing explicit micro-event replay with macro-step inference, this formulation provides a path toward substantially faster prediction of long-term materials evolution while preserving structural validity and time semantics. Across five atomic systems, spanning Cu-rich RPV steel irradiation aging, Cu-Zr metallic glass, and Li$_3$N-based anti-perovskite solid electrolyte, macro-step inference delivers a speed up of $10^3$ to $10^4$ times over event-by-event simulation.
Related papers
- Atom-JEPA: Joint-Embedding Predictive Architecture for 3D Atomistic Systems
- Thermodynamically Consistent Machine Learning for Fracture Prediction in Heterogeneous Materials
- Generative World Models Enable Predictive Control of Laser Melt Pool Dynamics
- Stable and Counterfactually Robust Physical World Models from Imposed Structure and Learned Physics
- Certified Access: Test-Time Compute on Conservative Physics-Structured AI Memory
Publication record
Bibliographic evidence describes where and when a record appeared. It does not validate scientific claims.
First public release: 2026-10-08 · Indexed by Pub-AI: 2026-10-09 UTC
AtomWorld-Mirror: Macro-Step World Modeling of Critical Evolution Backbones for Materials Dynamics
Preprint posted: 2026-10-08 · Original record ↗
- arxiv
- 2610.11527
Explicit source versions (1)
- v2 · 2026-10-10
- Explicit source version detected · v2
Source date: 2026-10-10 · First detected: 2026-10-25 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · atom:id explicit version (official metadata)
- Preprint available
Source date: 2026-10-08 · First detected: 2026-10-25 UTC · Historical baseline
Source evidence (1)
- Official evidence ↗ · atom:published (official metadata)
Abstract · arxiv
Atomistic simulation is a fundamental tool for studying long-term materials evolution, from diffusion and defect dynamics to interfacial reactions and fracture. Yet conventional simulators typically advance at microscopic resolution, spending substantial computation on low-impact local updates before reaching structurally consequential states, an evolutionary-resolution bottleneck that limits long-horizon simulation. We propose AtomWorld-Mirror, a time-aware macro-step world model for the critical evolution backbone of atomic systems. For Step-Wise atomistic simulation, AtomWorld-Mirror distills short micro-event segments into physically reachable transitions between key states, jointly predicting sparse structural edits and accumulated physical time through latent macro-step dynamics. Local reachability, inventory conservation, and continuous-time consistency constrain each transition. By amortizing local atomic physics into a reusable latent macro model and replacing explicit micro-event replay with macro-step inference, this formulation provides a path toward substantially faster prediction of long-term materials evolution while preserving structural validity and time semantics. Across five atomic systems, spanning Cu-rich RPV steel irradiation aging, Cu-Zr metallic glass, and Li$_3$N-based anti-perovskite solid electrolyte, macro-step inference delivers a speed up of $10^3$ to $10^4$ times over event-by-event simulation.