80264 research outputs found
Sort by
Hassan in Egypt [Electronic Version]
No full textContents: HASSAN IN EGYPT CHAPTER I HIGH NILE-- CHAPTER II FATHER AND SON -- CHAPTER III “THE OASIS OF ROSES” -- CHAPTER IV THE LITTLE RED CAMEL
-- CHAPTER V A GREEN AND RED DOOR -- CHAPTER VI MA'AS SALAMA -- CHAPTER VII AN UNLUCKY DAY -- CHAPTER VIII HASSAN IS LOST -- CHAPTER IX ON THE NILE -- CHAPTER X RAMESES THE GREAT -- CHAPTER XI FIVE THOUSAND YEARS AGO -- CHAPTER XII PICTURE WRITING -- CHAPTER XIII THE POTTER'S BOY -- CHAPTER XIV HASSAN'S NEW TUNIC -- CHAPTER XV THE GREAT DAM -- CHAPTER XVI A MIRAGE IN THE DESERT -- CHAPTER XVII PYRAMIDS OF EGYPT -- PRONOUNCING VOCABULARY AND DICTIONAR
Asian American Community Study: Discrimination and Anti-Asian Sentiment in the Houston Area (Survey Snapshot)
No full textAccording to Pew Research Center, nearly 6 in 10 Asian residents across the U.S. report experiencing discrimination because of their race, ethnicity, or perceived nationality in their lifetimes. Rates of experiencing discrimination differed across ethnicities. The Houston area is home to more than 676,000 Asian residents, made up of individuals and families from many different ethnicities, giving it one of the largest and most diverse populations of Asian communities in the country. The Asian American Community Study (AACS) asked Asian residents to report on experiences of discrimination they have had in the past year related to their nationality as well as their religion, gender, and age. Residents were also asked if they had experienced any anti-Asian sentiments or symbols. Findings: About 4 in 10 Asian residents in the Houston area reported experiences with discrimination in the past year; prevalence of experiences with discrimination was similar across most Asian ethnicities, with slightly lower rates among Vietnamese and Other Asian respondents and higher rates among those identifying with two or more races; more than 1 in 5 Asian young adults reported that they and their friends or family had experienced discrimination in the past year; compared to other experiences, discrimination on the basis of national origin was more often perpetrated by a stranger; nearly 4 in 10 Asian residents reported experiences with anti-Asian sentiments or symbols in the Houston area, with some differences across ethnicities; and almost half of Asian young adults experienced anti-Asian sentiments or symbols directly, had family or friends experience them, or both
Convergence Results and a New Preconditioner for Spectral Collocation in Time
No full textSpectral collocation methods provide a systematic construction for the approximate solution of ordinary differential equations (ODEs) of arbitrarily high order. These methods approximate the solution with a piecewise polynomial, which is determined by requiring the residual of the ODE to vanish at collocation points. This thesis presents three algebraically equivalent forms of the collocation method corresponding to different choices of polynomial bases. The convergence of global collocation for linear problems is analyzed from the viewpoint of projection methods, in which the projection operator represents interpolation by polynomials. This analysis is extended to nonlinear problems using the Kantorovich Theorem. Finally, a new preconditioner is presented that facilitates the efficient implementation of Chebyshev collocation methods. Numerical experiments demonstrate that the solution time of preconditioned spectral collocation behaves like O(K log K), where K is the number of collocation points, allowing for solves with over a million points
Optimizing Degree-of-Control in Reconfigurable Intelligent Surfaces
No full textIn each generation of wireless networks, there has been a dramatic increase in the available degrees of freedom (DoF). However, electronics to tap those degrees of freedom continue to lag as we add new spectral bands, e.g., we have limited Degrees of Control (DoC) to tap increasing DoF. In this thesis, we studied pattern-reconfigurable transmissive Reconfigurable Intelligent Surface (RIS)-assisted to increase degrees of control for a multiuser communication system. We jointly optimize the digital precoder, phase, and pattern parameters on the RIS and propose an alternating optimization method that combines conventional approaches with Deep Reinforcement Learning (DRL) algorithms. We employ our proposed approach for bit allocation between phase and pattern parameters under DoC constraints. Our findings highlight that phase control should be prioritized and also identify the regimes where pattern reconfigurability provides additional capacity gains
Reliable Medical LLM and Vision-Language RAG through Multi-Agent Orchestration and Single-Step Preference Alignment
No full textMedical RAG systems and long-context models like Med-PaLM face distinct yet
interconnected challenges in processing complex medical information. While RAG
systems struggle with hallucinations due to noisy retrievals and incomplete fact verification, long-context models, despite their ability to process extended inputs, suffer from attention dilution and context retention issues. Current BioNLP RAG systems have particularly overlooked the critical balance between retrieved context and parametric knowledge, often leading to hallucinations from over-reliance on retrieved information. Our HALO-MMedRAG framework addresses these challenges through a comprehensive multi-agent architecture. The system’s effectiveness stems from four innovative components: query-intent parser agent, multi-query generation agent, coarse retrieval agent, fine-grained hallucination aware retrieval agent with a perplexity and NLL based hybrid scoring for the chunks, generation agent, a light-weight fact-verification agent and an orchestrator agent that manages a CoT reasoning debate among 3 agents to provide the final hallucination free response, grounded in factuality.
The notion of Retrieval Augmented Generation in the context of Multimodal Medical LLMs has not been given due consideration from the lens of hallucination mitigation. Further, the existing approaches have been limited in their coverage of the medical domains, often limited to X-Ray. Medical Multimodal LLMs, when utilized for Multi-Modal Retrieval Augmented Generation, face critical challenges in maintaining factual accuracy while integrating complex visual and textual information. Our innovative approach addresses these challenges through a unified Triple Preference Optimization framework with three-stage preference dataset curation, focusing on cross-modal alignment, retrieval balance, and a dual staged visual feedback agent. Unlike existing solutions, our method employs a single-step optimization process that simultaneously handles multiple aspects of alignment while maintaining computational efficiency. Through careful curation of preference datasets that capture different levels of alignment quality, combined with a visual feedback agent for precise visual grounding to provide visual prompting for the Vision Language Model to improve its response, our approach significantly reduces hallucinations and improves medical response accuracy. Extensive evaluation across diverse medical domains, including radiology, ophthalmology, pathology, magnetic resonance imaging and CT scan demonstrates superior performance compared to the existing multimodal medical RAG methods, making our solution titled Align-MedRAG-VL, both practical and reliable for real-world medical applications where hallucination mitigation is paramount