Get Started
Research questionFor a known tensor-network graph, which structural parameters control MPS/TTN overhead and tomography complexity?The same tensor-network state can require different resources when represented as an MPS or TTN, depending on the structure of its underlying graph. That structure also affects how much data and computation are needed to learn the state, including when the input is not exactly representable by the chosen network.
Research Paper
Technology
Latest papersRecent research connected to this question, newest first.Parameterised graph theory for tensor networks: entanglement rerouting, structural simplification, and agnostic tomographyThe source studies tensor networks on arbitrary known graphs. It gives bounds involving cutwidth, tree-cutwidth, degree, treewidth, and learning complexity for bond-dimension overhead, TTN grouped-subsystem dimensions, and realisable tomography. Its agnostic setting considers approximating an arbitrary input state by a pure tensor-network state with a specified graph and bond dimension.research paper · Sep 3, 2026
Related questions
How can full-graph visibility improve fusion and hardware utilization during Transformer training?How can local loss-landscape geometry reveal algorithmic structure in learned Turing-machine computations?When do tied or untied attention parameterizations enable weak recovery under stochastic training?How can graph topology inference from node observations be unified with generating new graph instances?
Home
Topics
Search
Library