Geometric Characterisation and Structured Trajectory Surrogates for Clinical Dataset Condensation

• URL: http://arxiv.org/abs/2604.21638v1
• Date: Thu, 23 Apr 2026 12:53:46 GMT
• Title: Geometric Characterisation and Structured Trajectory Surrogates for Clinical Dataset Condensation
• Abstract summary: Trajectory matching (TM) is a widely used condensation approach that supervises synthetic data using changes in model parameters observed during training on real data.<n>We show that a fixed synthetic dataset can only reproduce a limited span of such training-induced parameter changes.<n>Motivated by this mismatch, we propose Bezier Trajectory Matching (BTM), which replaces SGD trajectories with quadratic Bezier trajectory surrogates.
• Score: 1.6261700844888172
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Dataset condensation constructs compact synthetic datasets that retain the training utility of large real-world datasets, enabling efficient model development and potentially supporting downstream research in governed domains such as healthcare. Trajectory matching (TM) is a widely used condensation approach that supervises synthetic data using changes in model parameters observed during training on real data, yet the structure of this supervision signal remains poorly understood. In this paper, we provide a geometric characterisation of trajectory matching, showing that a fixed synthetic dataset can only reproduce a limited span of such training-induced parameter changes. When the resulting supervision signal is spectrally broad, this creates a conditional representability bottleneck. Motivated by this mismatch, we propose Bezier Trajectory Matching (BTM), which replaces SGD trajectories with quadratic Bezier trajectory surrogates between initial and final model states. These surrogates are optimised to reduce average loss along the path while replacing broad SGD-derived supervision with a more structured, lower-rank signal that is better aligned with the optimisation constraints of a fixed synthetic dataset, and they substantially reduce trajectory storage. Experiments on five clinical datasets demonstrate that BTM consistently matches or improves upon standard trajectory matching, with the largest gains in low-prevalence and low-synthetic-budget settings. These results indicate that effective trajectory matching depends on structuring the supervision signal rather than reproducing stochastic optimisation paths.

Related papers

• Preference Data Selection for Mitigating the Alignment Tax in Large Language Models [52.18310687523229]
  BALIGN is a balanced data selection strategy that mitigates catastrophic forgetting while optimizing alignment efficacy.<n>By aggregating three key data-centric features into a unified composite risk score, BALIGN systematically filters out high-risk preference samples.<n>Experiments on standard human preference datasets demonstrate that BALIGN strongly preserves foundational capabilities without compromising alignment gains.
  arXiv  Detail & Related papers  (2026-08-25T07:58:42Z)
• Semi-Supervised Conditional Generative Learning through Stochastic Interpolation and Sufficient Representations [12.299177496689621]
  We propose a semi-supervised framework combining conditional intrinsic with low-dimensional latent representations.<n>RepG decomposes generation into two stages: label-dependent latent sampling and high-dimensional reconstruction.
  arXiv  Detail & Related papers  (2026-07-18T09:21:53Z)
• Supervised Distributional Reduction via Optimal Transport and Dependence Maximization [1.0499611180329804]
  We propose Supervised Distribution Reduction (SDR), an algorithm for learning target-aware representations.<n>SDR builds on the Fused Gromov-Wasserstein (FGW) objective to align the relational structure of the input distribution with a set of representative points.<n>SDR naturally induces a data-dependent, non-stationary geometry that can be leveraged for settings such as Gaussian Process (GP) modelling.
  arXiv  Detail & Related papers  (2026-05-26T19:38:20Z)
• Beyond Point-Wise Matching: Structural Representation Alignment for Accelerating Diffusion Transformers [93.3976834364707]
  We propose sREPA, a structural REPresentation Alignment framework to enforce consistency in the relational geometry of feature maps.<n>By encouraging the model to internalize holistic spatial layouts and structural correlations from pre-trained features, sREPA achieves faster and more stable convergence.
  arXiv  Detail & Related papers  (2026-05-16T12:01:04Z)
• TopoCurate:Modeling Interaction Topology for Tool-Use Agent Training [53.93696896939915]
  Training tool-use agents typically rely on Supervised Fine-Tuning (SFT) on successful trajectories and Reinforcement Learning (RL) on pass-rate-selected tasks.<n>We propose TopoCurate, an interaction-aware framework that projects multi-trial rollouts from the same task into a unified semantic quotient topology.<n>TopoCurate achieves consistent gains of 4.2% (SFT) and 6.9% (RL) over state-of-the-art baselines.
  arXiv  Detail & Related papers  (2026-03-02T10:38:54Z)
• GeodesicNVS: Probability Density Geodesic Flow Matching for Novel View Synthesis [54.39598154430305]
  We propose a Data-to-Data Flow Matching framework that learns deterministic transformations directly between paired views.<n>PDG-FM constrains flow trajectories using geodesic interpolants derived from probability density metrics of pretrained diffusion models.<n>These results highlight the advantages of incorporating data-dependent geometric regularization into deterministic flow matching for consistent novel view generation.
  arXiv  Detail & Related papers  (2026-03-01T09:30:11Z)
• Silent Inconsistency in Data-Parallel Full Fine-Tuning: Diagnosing Worker-Level Optimization Misalignment [27.352639822596146]
  Cross-worker divergence in losses and gradients can remain invisible under conventional monitoring signals.<n>We propose a model-agnostic diagnostic framework that quantifies worker-level consistency using training signals readily available in standard pipelines.
  arXiv  Detail & Related papers  (2026-02-16T04:42:30Z)
• Consolidation or Adaptation? PRISM: Disentangling SFT and RL Data via Gradient Concentration [56.074760766965085]
  PRISM achieves a dynamics-aware framework that arbitrates data based on its degree of cognitive conflict with the model's existing knowledge.<n>Our findings suggest that disentangling data based on internal optimization regimes is crucial for scalable and robust agent alignment.
  arXiv  Detail & Related papers  (2026-01-12T05:43:20Z)
• Improving Clinical Dataset Condensation with Mode Connectivity-based Trajectory Surrogates [15.665823714894605]
  State-of-the-art datasets condensation (DC) enables the creation of privacy-preserving synthetic datasets that can match the utility of real patient records.<n>DC methods supervise synthetic data by aligning the training dynamics of models trained on real and those trained on synthetic data.<n>We address these limitations by replacing full SGD trajectories with smooth, low-loss parametric surrogates.<n>These mode-connected paths provide noise-free, low-curvature supervision signals that stabilise gradients, accelerate convergence, and eliminate the need for dense trajectory storage.
  arXiv  Detail & Related papers  (2025-10-07T11:22:27Z)
• Bridging the Synthetic-Real Gap: Supervised Domain Adaptation for Robust Spacecraft 6-DoF Pose Estimation [13.83897333268682]
  Spacecraft Pose Estimation is a fundamental capability for autonomous space operations such as rendezvous, docking, and in-orbit docking.<n>Existing domain adaptation approaches aim to mitigate this issue but often underperform when a modest number of labeled target samples are available.<n>We propose the first Supervised Domain Adaptation (SDA) framework tailored for SPE keypoint regression.
  arXiv  Detail & Related papers  (2025-09-17T08:03:05Z)
• Hierarchical Features Matter: A Deep Exploration of Progressive Parameterization Method for Dataset Distillation [44.03611131165989]
  We propose a novel generative parameterization method dubbed Hierarchical generative Distillation (H-PD)<n>The proposed H-PD achieves a significant performance improvement under various settings with equivalent time consumption.<n>It even surpasses current generative distillation using diffusion models under extreme compression ratios IPC=1 and IPC=10.
  arXiv  Detail & Related papers  (2024-06-09T09:15:54Z)
• CAFE: Learning to Condense Dataset by Aligning Features [72.99394941348757]
  We propose a novel scheme to Condense dataset by Aligning FEatures (CAFE) At the heart of our approach is an effective strategy to align features from the real and synthetic data across various scales. We validate the proposed CAFE across various datasets, and demonstrate that it generally outperforms the state of the art.
  arXiv  Detail & Related papers  (2022-03-03T05:58:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.