2787 papers
This paper presents a privacy-preserving framework that generates clinically accurate synthetic data to train a DistilBERT model for detecting coercive control, effectively overcoming the "Forensic Data Gap" by achieving high performance on real-world legal and social media benchmarks without compromising user privacy.
This paper proposes a novel "Symmetric Cartesian Collapse" method that simulates non-Clifford quantum circuits in linear time and storage by modeling quantum systems as single discrete trajectories rather than dense matrices, theoretically enabling the simulation of over a thousand qubits on consumer hardware.
This paper demonstrates that in continuous thought models, readout-based interpretability often fails to reflect underlying computation because the latent states required to change the model's answer are significantly larger than those needed to alter the projected readout, necessitating dose-response curves rather than single-magnitude interventions for accurate causal analysis.
This paper proposes a decentralized, AI-powered Zero-Trust identity management system that enhances security in fintech ecosystems by replacing vulnerable OTPs with deepfake-resistant multimodal biometrics and predictive cyber defense, achieving 99.2% threat detection accuracy.
This systematic review and meta-analysis demonstrates that multi-agent collaboration systems significantly outperform single large language models in medical question answering, despite critical gaps in reporting transparency, geographic concentration in China and the US, and variable cost-effectiveness.
This paper introduces BCSDBT-Seg, an open-access dataset featuring expert-annotated 3D voxel-level lesion segmentation masks and structured BI-RADS reports for 396 digital breast tomosynthesis volumes from 201 biopsy-confirmed patients, designed to advance automated segmentation, radiomics, and deep learning research in breast cancer screening.
JUDICATA is a model-agnostic, agentic neuro-symbolic architecture that integrates adaptive retrieval, logical reasoning chains, adversarial defense, and Toulmin-based structured output to significantly reduce hallucinations and enhance the reliability, security, and procedural contestability of automated legal reasoning.
This paper evaluates the robustness of the REFORGE-based forensic system against platform-induced distortions across four social networks and four datasets, finding that while binary tamper decisions remain stable, confidence scores and mask geometries exhibit measurable drift, thereby motivating the adoption of OSN-consistency regularization and platform-conditioned normalization for enhanced detection reliability.
This paper proposes PatchMamS2S, a lightweight Mamba-2-based multivariate time series forecasting model tailored for the power domain that combines patch mining, physical prior-guided weight fusion, and an efficient Seq2Seq architecture to overcome the accuracy-resource trade-offs of existing universal models.
BioKG Explorer is a full-stack, open-source web application that enables interactive, no-code visual analytics and hypothesis generation over the large-scale PrimeKG biomedical knowledge graph by combining Neo4j-based graph traversal, disease-specific analytics, and large-language-model path explanations within a browser-based interface.