AI
-
Accelerating LLM Inference via Dynamic KV Cache Placement in Heterogeneous Memory System
Source: arXiv AI PapersRead more: Accelerating LLM Inference via Dynamic KV Cache Placement in Heterogeneous Memory SystemThis work addresses the challenge of memory bandwidth constraints in Large Language Model (LLM) inference caused by frequent key-value (KV)…
-
The Role of AI in Facilitating Interdisciplinary Collaboration: Evidence from AlphaFold
Source: arXiv AI PapersRead more: The Role of AI in Facilitating Interdisciplinary Collaboration: Evidence from AlphaFoldThis study investigates the role of AI, specifically AlphaFold, in promoting interdisciplinary collaboration among scientists. The findings reveal that AlphaFold…
-
Uncertainty-Aware Learning Policy for Reliable Pulmonary Nodule Detection on Chest X-Ray
Source: arXiv AI PapersRead more: Uncertainty-Aware Learning Policy for Reliable Pulmonary Nodule Detection on Chest X-RayThis study proposes an Uncertainty-Aware Learning Policy to improve pulmonary nodule detection on chest X-rays by integrating physicians’ background knowledge…
-
ViTAD: Timing Violation-Aware Debugging of RTL Code using Large Language Models
Source: arXiv AI PapersRead more: ViTAD: Timing Violation-Aware Debugging of RTL Code using Large Language ModelsViTAD is a novel method for debugging timing violations in RTL code using large language models (LLMs). It automates root…
-
Hierarchical Conformal Classification
Source: arXiv AI PapersRead more: Hierarchical Conformal ClassificationHierarchical Conformal Classification (HCC) extends conformal prediction by incorporating class hierarchies into prediction sets, maintaining reliable uncertainty quantification with coverage…
-
GaitCrafter: Diffusion Model for Biometric Preserving Gait Synthesis
Source: arXiv AI PapersRead more: GaitCrafter: Diffusion Model for Biometric Preserving Gait SynthesisGaitCrafter is a diffusion-based framework designed to synthesize realistic and identity-preserving gait sequences from silhouette data, addressing the scarcity of…
-
Overcoming Latency Bottlenecks in On-Device Speech Translation: A Cascaded Approach with Alignment-Based Streaming MT
Source: arXiv AI PapersRead more: Overcoming Latency Bottlenecks in On-Device Speech Translation: A Cascaded Approach with Alignment-Based Streaming MTThis paper addresses latency challenges in real-time on-device speech translation by proposing a cascaded approach that integrates Automatic Speech Recognition…
-
Whispering Context: Distilling Syntax and Semantics for Long Speech Transcripts
Source: arXiv AI PapersRead more: Whispering Context: Distilling Syntax and Semantics for Long Speech TranscriptsThis work addresses the challenge of maintaining syntactic and semantic accuracy in long speech transcripts by enhancing ASR systems. The…
-
AdaptJobRec: Enhancing Conversational Career Recommendation through an LLM-Powered Agentic System
Source: arXiv AI PapersRead more: AdaptJobRec: Enhancing Conversational Career Recommendation through an LLM-Powered Agentic SystemAdaptJobRec is a novel conversational job recommendation system that uses an autonomous agent to balance complex query handling and response…
-
ALIGN: Word Association Learning for Cross-Cultural Generalization in Large Language Models
Source: arXiv AI PapersRead more: ALIGN: Word Association Learning for Cross-Cultural Generalization in Large Language ModelsThe ALIGN study introduces a method to improve cultural alignment in large language models by fine-tuning them on native speakers’…