Bridging the Gap Between Plausibility and Admissibility: Constraint-Aware Flow Maps for Dynamic Graph Systems

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Michael Romei de Socio, Gian Luca Pozzato, Alessio Merlo
• Venue: arXiv: cs.AI
• Published: 2026-07-23
• Collected: 2026-07-25
• URL: http://arxiv.org/abs/2607.21421v1
• PDF: https://arxiv.org/pdf/2607.21421v1
• DOI: UNKNOWN

Plain-Language Brief

A new model is proposed to improve the accuracy of generative trajectory modeling in dynamic graph-structured systems by applying hard filtering, soft weighting, or projection-based repair using post-sampling symbolic constraints. The study evaluates this model on two synthetic regimes and finds that while statistical plausibility and structural admissibility are distinct reliability properties, the new model shows promising results in reducing invalid trajectories as graph complexity increases.

What the Document Says

• The text discusses a study investigating whether post-sampling symbolic constraints can improve the reliability of generative trajectory modeling for dynamic graph-structured systems.
• A conditional diffusion model generates future graph-state trajectories from partial observations, and an external symbolic layer applies hard filtering, soft weighting, or projection-based repair.
• The study evaluates the framework on two synthetic regimes: a compact graph and a medium-complexity dependency graph.
• Metrics for structural validity, sample efficiency, diversity, robustness, and calibration are used.
• In the compact regime, the model produces an invalid probability mass of 0.002996, indicating an almost entirely admissible trajectory manifold.
• Under the same architecture and training protocol, invalid mass increases to 0.155929 in the medium-complexity regime.
• Hard filtering removes all invalid retained trajectories while preserving 84.4% of generated samples.
• Soft weighting preserves effective sample size but yields only limited validity gains.
• Dependency constraints account for nearly all observed inadmissibility in a family-level analysis.
• The results suggest that statistical plausibility and structural admissibility are distinct reliability properties, and symbolic constraint handling becomes more valuable as graph-structural complexity increases.

Key Concepts Mentioned

• Generative models
• Decision-making under uncertainty
• Dynamic graph-structured systems
• Conditional diffusion model
• Symbolic constraints
• Hard filtering
• Soft weighting
• Dependency graph
• Structural validity
• Sample efficiency
• Diversity
• Robustness
• Calibration

Claims Made by the Document

• Post-sampling symbolic constraints can improve the reliability of generative trajectory modeling.
• Statistical plausibility does not ensure structural feasibility.

Evidence Used

• The study investigates whether post-sampling symbolic constraints can improve the reliability of generative trajectory modeling.
• Increase in invalid probability mass from compact to medium-complexity regime.
• Results showing that hard filtering and soft weighting have different impacts on validity.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Generative models can support decision-making under uncertainty by producing ensembles of plausible future system trajectories, but statistical plausibility does not ensure structural feasibility. This study investigates whether post-sampling symbolic constraints can improve the reliability of generative trajectory modeling in dynamic graph-structured systems. A conditional diffusion model generates future graph-state trajectories from partial observations, while an external symbolic layer applies hard filtering, soft weighting, or projection-based repair. The framework is evaluated on two controlled synthetic regimes: a compact graph and a medium-complexity dependency graph, using metrics for structural validity, sample efficiency, diversity, robustness, and calibration. In the compact regime, the model produces an invalid probability mass of 0.002996, indicating an almost entirely admissible trajectory manifold. Under the same architecture and training protocol, invalid mass increases to 0.155929 in the medium-complexity regime. Hard filtering removes all invalid retained trajectories while preserving 84.4% of generated samples, whereas soft weighting preserves effective sample size but yields only limited validity gains. Family-level analysis shows that dependency constraints account for nearly all observed inadmissibility. These results indicate that statistical plausibility and structural admissibility are distinct reliability properties and that symbolic constraint handling becomes more valuable as graph-structural complexity increases.