@animaanandkumar - AI Dynamics

(2)The Loss function defined in @UCIrvine paper also appears to become infinite while we ensure the loss is finite and results in valid Langevin dynamics.

→ View original post on X — @animaanandkumar

21 February 2023

Diffusion Models in Function Spaces: Key Differences from Prior Work

By

@animaanandkumar

–

21 February 2023 2h25

Yes, we are aware of paper on diffusion models in function spaces from @UCIrvine There are two main differences from our work: (1) We assume data in function space vs. other work: data discretized with IID Gaussian noise. Latter not well-defined as resolution increases. See Fig

→ View original post on X — @animaanandkumar

21 February 2023

Joint Research Collaboration with Multiple Institutions and Partners

By

@animaanandkumar

–

17 February 2023 2h57

Joint work with @shrimai_ @rmichaelalvarez @caltech @nvidia

→ View original post on X — @animaanandkumar

17 February 2023

Automated Testing for Social Bias in Large Language Models

By

@animaanandkumar

–

17 February 2023 2h56

Testing Language Models at scale for social bias is challenging. We build automated testing: test sentences are automatically generated given the bias dimensions. This allows getting statistically meaningful measures for bias as opposed to a small number of hand-written templates

→ View original post on X — @animaanandkumar

17 February 2023

MXNet’s Community Engagement Issues and Documentation Importance

By

@animaanandkumar

–

16 February 2023 21h55

Mxnet suffered from a lack of engagement with the community. I saw it first hand. You can take users for granted. Ease of use and clear documentation goes a long way.

→ View original post on X — @animaanandkumar

16 February 2023