3020 papers
This study employs a candidate-centric framework to demonstrate that trust is the strongest predictor of job applicants' intention to use AI-driven talent acquisition systems, significantly mediating the impact of ethical concerns like fairness, transparency, and privacy within the Indian context.
This paper presents a resource-efficient fine-tuning methodology using Low-Rank Adaptation (LoRA) and quantization to successfully deploy large Vision-Language-Action models on affordable, consumer-grade robotic platforms, demonstrating effective real-world manipulation performance with limited data.
This paper presents an LSTM-based deep learning framework that utilizes transfer learning on encrypted network flow features to detect ransomware without payload decryption, demonstrating the critical necessity of 12–18 month retraining cycles to counteract temporal model drift while achieving high detection accuracy through a confidence-based ensemble system integrated with Wazuh SIEM.
This systematic review of literature from 2020 to 2026 evaluates the impact of immersive virtual reality on diverse educational outcomes, finding that while VR significantly enhances engagement, retention, and conceptual understanding, its widespread adoption is currently hindered by barriers such as cost, cybersickness, and insufficient teacher training, thereby outlining a roadmap for future research in adaptive learning, analytics, and ethical implementation.
This paper establishes a formal correspondence between predictive coding network dynamics and Pearl's do-calculus, demonstrating through theoretical derivation and empirical validation on clinical benchmarks that specific architectural interventions enable these networks to accurately perform causal reasoning and outperform standard associational models in estimating interventional effects.
This paper proposes the AI-Assisted Music Creation Framework (AAMCF), a conceptual five-layer architecture that leverages neural pitch correction, generative harmonization, and adaptive feedback to enable individuals with amusia to create and perceive music, while outlining its theoretical foundations, evaluation metrics, and ethical considerations for future empirical validation.
This paper introduces a post-processing MLIR analysis pass for the snn-mlir compiler that generates machine-checkable worst-case latency proofs and memory-safety certificates, thereby enabling the IEC 62304 Class B compliant deployment of spiking neural networks for safety-critical edge medical devices like seizure detectors.
This paper identifies and validates a previously unarticulated "max_cluster = 1" condition, requiring pairwise distinctness of all RoPE frequencies, demonstrating that violating this implicit constraint at inference time causes catastrophic performance degradation in large language models while proving that standard geometric schedules inherently satisfy it due to the transcendental nature of their logarithmic bases.
This paper proposes the Dynamic Balancing Hybrid Merkle Tree (DBHMT), an optimized data structure for IOTA Tangle-based Healthcare 4.0 that employs a dynamic pruning mechanism to achieve a fourfold reduction in storage costs while maintaining security for resource-constrained IoT environments.
This systematic review synthesizes 15 empirical studies from 2022 to 2026, finding that while retrieval-augmented generation (RAG) configurations generally outperform baselines in mental health applications, the field is hindered by heterogeneous evaluation methods, sparse safety reporting, and a lack of multimodal data integration.