skip to main content


Search for: All records

Creators/Authors contains: "Zhu, Y."

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. We consider concept generalization at a large scale in the diverse and natural visual spectrum. Established computational modes (i.e., rule-based or similarity-based) are primarily studied isolated and focus on confined and abstract problem spaces. In this work, we study these two modes when the problem space scales up, and the complexity of concepts becomes diverse. Specifically, at the representational level, we seek to answer how the complexity varies when a visual concept is mapped to the representation space. Prior psychology literature has shown that two types of complexities (i.e., subjective complexity and visual complexity) build an inverted-U relation. Leveraging the Representativeness of Attribute (RoA), we computationally confirm the following observation: Models use attributes with high RoA to describe visual concepts, and the description length falls in an inverted-U relation with the increment in visual complexity. At the computational level, we aim to answer how the complexity of representation affects the shift between the rule- and similarity-based generalization. We hypothesize that category-conditioned visual modeling estimates the co-occurrence frequency between visual and categorical attributes, thus potentially serving as the prior for the natural visual world. Experimental results show that representations with relatively high subjective complexity out-perform those with relatively low subjective complexity in the rule-based generalization, while the trend is the opposite in the similarity-based generalization. 
    more » « less
    Free, publicly-accessible full text available October 1, 2024
  2. Abstract

    We report the results of thermodynamic measurements in external magnetic field of the cubic Ce-based cage compounds CeT2Cd20(T= Ni,Pd). Our analysis of the heat-capacity data shows that the Γ7doublet is the ground state multiplet of the Ce3+ions. Consequently, for the Γ7doublet it can be theoretically shown that the Ruderman–Kittel–Kasuya–Yosida interaction between the localized Ce moments mediated by the conduction electrons, must vanish at temperatures much lower than the energy separating the ground state doublet from the first excited Γ8quartet. Our findings provide an insight as to why no long range order has been observed in these compounds down to temperatures in the milliKelvin range.

     
    more » « less
  3. Inspired by humans’ exceptional ability to master arithmetic and generalize to new problems, we present a new dataset, Handwritten arithmetic with INTegers (HINT), to examine machines’ capability of learning generalizable concepts at three levels: perception, syntax, and semantics. In HINT, machines are tasked with learning how concepts are perceived from raw signals such as images (i.e., perception), how multiple concepts are structurally combined to form a valid expression (i.e., syntax), and how concepts are realized to afford various reasoning tasks (i.e., semantics), all in a weakly supervised manner. Focusing on systematic generalization, we carefully design a five-fold test set to evaluate both the interpolation and the extrapolation of learned concepts w.r.t. the three levels. Further, we design a few-shot learning split to determine whether or not models can rapidly learn new concepts and generalize them to more complex scenarios. To comprehend existing models’ limitations, we undertake extensive experiments with various sequence-to-sequence models, including RNNs, Transformers, and GPT-3 (with the chain of thought prompting). The results indicate that current models struggle to extrapolate to long-range syntactic dependency and semantics. Models exhibit a considerable gap toward human-level generalization when evaluated with new concepts in a few-shot setting. Moreover, we discover that it is infeasible to solve HINT by merely scaling up the dataset and the model size; this strategy contributes little to the extrapolation of syntax and semantics. Finally, in zero-shot GPT-3 experiments, the chain of thought prompting exhibits impressive results and significantly boosts the test accuracy. We believe the HINT dataset and the experimental findings are of great interest to the learning community on systematic generalization. 
    more » « less
    Free, publicly-accessible full text available May 1, 2024
  4. Free, publicly-accessible full text available May 1, 2024
  5. Abstract

    Our basic knowledge of downward positive lightning leaders is incomplete due to their rarity and limited ability of VHF mapping systems to image positive streamers. Here, using high‐speed optical records and wideband electric field and magnetic field derivative signatures, we examine in detail the development of a descending positive leader, which extended intermittently via alternating branching at altitudes of 4.2 to 1.9 km and involved luminosity transients separated by millisecond‐scale quiet intervals. We show that the transients (a) are mostly initiated in previously created but already decayed branches, at a distance of the order of 100 m above the branch lower extremity, (b) extend bidirectionally with negative charge moving up, (c) establish a temporary (1 ms or so) steady‐current connection to the negative part of the overall bidirectional leader tree, and (d) exhibit brightening accompanied by new breakdowns at the positive leader end. One of the transients unexpectedly resulted in a negative cloud‐to‐ground discharge. Both positive and negative ends of the transients extended at speeds of 106–107 m/s, while the overall positive leader extension speed was as low as 103–104 m/s. Wideband electric field signatures of the transients were similar to K‐changes, with their millisecond‐ and microsecond‐scale features being associated with the steady current and new breakdowns, respectively. For transients with both ends visible in our optical records, charge transfers and average currents were estimated to be typically a few hundreds of millicoulombs and some hundreds of amperes, respectively.

     
    more » « less