skip to main content

Explore Research Products in the NSF-PAR

  • Home
  • Search Results
  • Page 1 of 18

Search for: All records

Creators/Authors contains: "Kong, D."

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. ; ; ; ; ; ; ( , International Conference on Machine Learning (ICML 2023))
    The capability to generate responses with diversity and faithfulness using factual knowledge is paramount for creating a human-like, trustworthy dialogue system. Common strategies either adopt a two-step paradigm, which optimizes knowledge selection and response generation separately and may overlook the inherent correlation between these two tasks, or leverage conditional variational method to jointly optimize knowledge selection and response generation by employing an inference network. In this paper, we present an end-to-end learning framework, termed Sequential Posterior Inference (SPI), capable of se- lecting knowledge and generating dialogues by approximately sampling from the posterior distribution. Unlike other methods, SPI does not require the inference network or assume a simple geometry of the posterior distribution. This straightforward and intuitive inference procedure of SPI directly queries the response generation model, allowing for accurate knowledge selection and generation of faithful responses. In addition to modeling contributions, our experimental results on two common dialogue datasets (Wizard of Wikipedia and Holl-E) demonstrate that SPI outperforms previous strong baselines according to both automatic and human evaluation metrics. 
    more » « less
    Free, publicly-accessible full text available October 1, 2024
  2. ; ; ; ( , Uncertainty in Artificial Intelligence (UAI, 2023))
    Generation of molecules with desired chemical and biological properties such as high drug-likeness, high binding affinity to target proteins, is critical for drug discovery. In this paper, we propose a probabilistic generative model to capture the joint distribution of molecules and their properties. Our model assumes an energy-based model (EBM) in the latent space. Conditional on the latent vector, the molecule and its properties are modeled by a molecule generation model and a property regression model respectively. To search for molecules with desired properties, we propose a sampling with gradual distribution shifting (SGDS) algorithm, so that after learning the model initially on the training data of existing molecules and their properties, the proposed algorithm gradually shifts the model distribution towards the region supported by molecules with desired values of properties. Our experiments show that our method achieves very strong performances on various molecule design tasks. 
    more » « less
    Free, publicly-accessible full text available August 1, 2024
  3. ; ; ; ; ; ; ; ; ; ; et al ( , Osteoarthritis and Cartilage)
    null (Ed.)
    Full Text Available 
  4. ; ; ; ; ; ; ( , Journal of Microscopy)
    Summary Lay Description

    Centrioles are microtubule‐based structures organised as ninefold symmetrical cylinders which are, in human cells, ∼500 nm long and ∼230 nm wide. Centrioles assemble dozens of proteins around them forming centrosomes, which nucleate microtubules and organise spindle poles in mitosis. Centrioles, in addition, assemble cilia and flagella, two critically important organelles for signalling and motility. Due to centriole small size, electron microscopy has been a major imaging technique for the analysis of their ultrastructural features. However, being technically demanding, electron microscopy it is not easily available to the researchers and it is rarely used to collect large datasets. Expansion microscopy is an emerging approach in which biological specimens are embedded in a swellable polymer and isotopically expanded several fold. Physical separation of cellular structures allows the analysis of, otherwise unresolvable, structures by conventional optical microscopes. We present an adaptation of expansion microscopy approach, specifically developed for a robust analysis of centrioles and cilia. Our protocol can be used for the analysis of centriole number, duplication status, length, localisation of various centrosomal components and ciliation from large populations of cultured adherent and nonadherent cells and multiciliated cultures. We validate the method against electron microscopy and superresolution microscopy and demonstrate that it can be used as an accessible and reliable alternative to electron microscopy.

     
    more » « less
  5. ; ; ; ; ; ; ; ; ; ; et al ( , Physical Review D)
  6. ; ; ; ; ; ; ; ; ; ; et al ( , Physical Review D)
  7. ; ; ; ; ; ; ; ; ; ; et al ( , Physical Review D)
  8. ; ; ; ; ; ; ; ; ; ; et al ( , Physical Review D)
  9. ; ; ; ; ; ; ; ; ; ; et al ( , Physical Review D)
  10. ; ; ; ; ; ; ; ; ; ; et al ( , Physical Review D)