A5058670437- Recommender Systems and Techniques
- Photoreceptor and optogenetics research
- Receptor Mechanisms and Signaling
- Neuroscience and Neuropharmacology Research
- Cancer-related molecular mechanisms research
- Infrared Thermography in Medicine
- Ovarian cancer diagnosis and treatment
- Ovarian function and disorders
- RNA modifications and cancer
- Machine Learning in Healthcare
- RNA Research and Splicing
- Advanced Chemical Sensor Technologies
- Multimedia Communication and Technology
- Information Retrieval and Search Behavior
- Video Analysis and Summarization
- Topic Modeling
- Optical Imaging and Spectroscopy Techniques
- Extracellular vesicles in disease
- Image Retrieval and Classification Techniques
Google (United States)
2024
Peking University
2023
Chinese Institute for Brain Research
2023
Huazhong University of Science and Technology
2013
University of Arizona
2009
Dopamine (DA) plays multiple roles in a wide range of physiological and pathological processes via vast network dopaminergic projections. To fully dissect the spatiotemporal dynamics DA release both dense sparsely innervated brain regions, we developed series green red fluorescent GPCR activation-based (GRAB
Sequential models are invaluable for powering personalized recommendation systems. In the context of short-form video (SFV) feeds, where user behavior history is typically longer, systems must be able to understand users' long-term interests. However, deploying large sequence extensive web-scale applications faces challenges due high serving cost. To address this, we propose an industrial framework designed efficiently models. Specifically, proposed infrastructure decouples model and main...
Randomly-hashed item ids are used ubiquitously in recommendation models. However, the learned representations from random hashing prevents generalization across similar items, causing problems of learning unseen and long-tail especially when corpus is large, power-law distributed, evolving dynamically. In this paper, we propose using content-derived features as a replacement for ids. We show that simply replacing ID with content-based embeddings can cause drop quality due to reduced...
Recent progress in large language models (LLMs) offers promising new approaches for recommendation system (RecSys) tasks. While the current state-of-the-art methods rely on fine-tuning LLMs to achieve optimal results, this process is costly and introduces significant engineering complexities. Conversely, that bypass use directly are less resource-intensive but often fail fully capture both semantic collaborative information, resulting sub-optimal performance compared their fine-tuned...