Task Embedding

Task Embedding is a representational encoding that captures the distinctive characteristics of a learning task within a shared latent space, enabling meta-learning systems to generalize across tasks by modeling task-specific inductive biases as compact vectorial signatures.

30

Publications

2.1K

Citations

120

Authors

61

Institutions

Publications per year

2017–2026

24

Authors

120

  1. XW
    Xin Wang

    Stony Brook University, United States

    2 publications · 125 citations

    2

    h-index

  2. FY
    Fisher Yu

    University of California, Berkeley, United States

    2 publications · 125 citations

    2

    h-index

  3. JE
    Joseph E. Gonzalez

    University of California, Berkeley, United States

    2 publications · 125 citations

    2

    h-index

  4. TD
    Trevor Darrell

    University of California, Berkeley, United States

    2 publications · 125 citations

    2

    h-index

  5. RW
    Ruth Wang

    University of California San Diego, United States

    2 publications · 125 citations

    2

    h-index

1–5 of 120

Venues

  1. arXiv (Cornell University)

    Repository · 5 publications · 119 citations

    5

    h-index

  2. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 3 publications · 60 citations

    3

    h-index

  3. eScholarship (California Digital Library)

    Repository · 1 publications · 21 citations

    1

    h-index

  4. Neurocomputing

    1 publications · 20 citations

    1

    h-index

  5. Behavior Therapy

    1 publications · 22 citations

    1

    h-index

1–5