Data Curation & Synthetic Data - Weekly Roundup

A physics-enhanced machine learning approach for the system identification of Linear Time-Varying systems under time-varying operating conditions in terms of fast-varying natural frequencies and damping ratios by combining a long short-term memory network with an Extended Kalman Filter is proposed.

This work generates a robust synthetic dataset of forward DTOF using exact Monte Carlo simulations at multiple source-detector distances and proposes a machine learning framework as an alternative approach to reconstruct the optical properties of a bilayered medium, benchmarking its efficiency and accuracy against model-based algorithms.

Three compact scratch-trained computer-vision models are benchmarked on the Notre Dame subset of LivDet-Iris 2017 under PAI-driven domain shift and environmental degradation to demonstrate deployment readiness under unknown PAIs.

The results suggest that gradient-based attribution can identify data responsible for subliminal learning in some settings, but that some approximations are more reliable than others.

A physics-enhanced machine learning approach for the system identification of Linear Time-Varying systems under time-varying operating conditions in terms of fast-varying natural frequencies and damping ratios by combining a long short-term memory network with an Extended Kalman Filter is proposed.

This work generates a robust synthetic dataset of forward DTOF using exact Monte Carlo simulations at multiple source-detector distances and proposes a machine learning framework as an alternative approach to reconstruct the optical properties of a bilayered medium, benchmarking its efficiency and accuracy against model-based algorithms.

Three compact scratch-trained computer-vision models are benchmarked on the Notre Dame subset of LivDet-Iris 2017 under PAI-driven domain shift and environmental degradation to demonstrate deployment readiness under unknown PAIs.

The results suggest that gradient-based attribution can identify data responsible for subliminal learning in some settings, but that some approximations are more reliable than others.

This research in progress paper investigates different approaches on how to combine JETs with ML-methods in a hybrid manner and presents specialized models to increase the detection performance and validity of anomaly detection results to improve audit efficiency.

This work highlights that diagnosing and factually augmenting data to long-tail deficits is essential for robust scientific retrieval, and presents REPAIR, a self-evolving data augmentation framework for scientific dense retrievers.

For the datasets and models considered, pixel-level image-fidelity metrics alone may not consistently capture the task-relevant realism of synthetic sonar observations and motivate the use of task-aware evaluation for synthetic sensor data intended for robotic perception.

MiST (Mid-trained Security Transformer), a suite of 8B and 32B models that achieve strong performance on public cybersecurity benchmarks are presented and it is shown that MiST provides a stronger initialization for downstream task-specific fine-tuning adaptation and reinforcement learning.

A framework for evaluating data-science agents on live, continuously updated data using executable ground truth and format-agnostic factoid scoring, which achieves a 29% improvement in the Matthews Correlation Coefficient and a 16% reduction in token consumption per test case.

This paper presents an end-to-end approach for generating context-specific large language model (LLM) benchmark datasets by combining expert input with synthetic data generation by introducing a schema eliciting key information about the goals, scope, and context of an evaluation task.

A matched ablation on unfiltered training data shows substantial benefits from filtering for all fine-tuned models, with the LLM-based models gaining roughly twice as much from filtering as the dedicated encoder-decoder MT system, supporting the view that for Belarusian MT one of the primary bottlenecks is data quality.

These findings provide an empirical characterization of recurring security patterns across the open-source onboard satellite software ecosystem, helping quantify their prevalence and prioritize areas that warrant the greatest security attention.