Resources
Papers, datasets, and work we're tracking.
A running library of papers and datasets relevant to our research directions — some we benchmark against directly, others are just worth reading. Grouped by the working group each is most relevant to; a Featured mark means it also shows up on that direction's card on the homepage.
Learned Video Representations
-
A comparison point we track for coordinate-based flow video representations.
- FFNeRV paper
NeRV-family baseline commonly benchmarked against. Link pending.
-
A hierarchical NeRV-family baseline we benchmark against.
- HNeRV paper
NeRV-family baseline commonly benchmarked against. Link pending.
- NeRF paper
Foundational neural field work, relevant background for structured latent representations. Link pending.
- NeRV paper
NeRV-family baseline commonly benchmarked against. Link pending.
- SIREN paper
Foundational implicit neural representation work, relevant background for structured latent representations. Link pending.
- UVG Dataset dataset Featured
The video dataset Nika is trained and evaluated on. Link pending.
Neural Operators
-
Large-scale physics simulation datasets, including the Rayleigh–Bénard convection data Niko benchmarks against.
Scientific Computing
-
A dataset of QI-like stellarator plasma boundaries and optimization benchmarks, relevant to our fusion work.
Neural Infrastructure
Nothing tracked here yet.