2661 papers
The ECN-BENCH paper presents a systems engineering audit of a multi-agent forecasting testbed, revealing that LLM-based probability extraction is highly sensitive to configuration choices and evaluator selection, thereby undermining the reliability of near-uniform forecasts and limiting the statistical significance of comparative performance results.
This paper introduces DiligenceProv, a novel benchmark and construction methodology using synthetic deal rooms with registered imperfections to evaluate and improve the verifiability, evidence retrieval, and definitional accuracy of large language models in high-stakes financial due diligence.
This paper proposes a robust, multi-layered audio encryption algorithm that combines the Lorenz chaotic system for initial key generation and XOR-based scrambling with a second layer utilizing the Elzaki transform and hyperbolic Maclaurin series expansion, demonstrating high security and efficiency through experimental validation and comparative performance analysis.
This paper proposes TADNet, a hierarchical encoder-decoder network that integrates a cascaded Swin-Transformer encoder with Transparency-Aware Attention and Structural Feature Enhancement modules to effectively address depth noise and geometric distortion in transparent object depth completion, achieving superior performance on the ClearGrasp and TransCG datasets.
This structured review critiques existing evaluations of tool-using LLM agents for failing to substantiate claims about preventing unauthorized resource effects, and proposes the Minimum Authorization-Assurance Profile (MAAP) as a novel, evidence-based framework to standardize the reporting of enforcement semantics, observation boundaries, and residual assumptions for transparent security comparison.
This paper proposes a Verification-First Agentic RAG framework that integrates deterministic validation agents to ensure regulatory accuracy and prevent hallucinations in Urban Cyber-Physical Systems, demonstrating superior performance in supersession detection and compliance reasoning compared to existing methods.
This cross-lingual agent-based audit of Turkish political discourse reveals that while prompt language minimally affects the political stance of generative AI responses, it significantly alters the evidentiary sources models rely on, with model identity and user persona exerting a stronger influence on the final output than the language of the prompt itself.
This study proposes an early-warning decision-support framework using ensemble learning models trained on the OULAD dataset and enhanced with SMOTE to effectively predict at-risk student performance at the first-quarter boundary, demonstrating that LightGBM achieves superior accuracy in identifying students needing timely academic intervention.
This paper proposes a novel framework for 6G edge networks that integrates privacy-preserving federated learning, Grey Wolf Optimizer-based trust model optimization, and a permissioned blockchain to achieve robust, resilient, and private trust management against adversarial attacks.
This paper demonstrates that LLM-based concept mapping systems can significantly reduce over-alignment errors on non-mappable terms by structuring candidate sets into comparable and excludable formats, a design principle that proves more effective than simply providing the correct answer or improving retrieval.