SciML / SciML/NeuralOperators.jl

Implement Latent Spectral Model (LSM)

Open
#121 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Julia
Stars
41
Forks
15
Avg merge
13h 14m
Merged PRs (30d)
12

Description

Summary

Implement LSM, which encodes input functions into a latent space, constructs an orthogonal spectral basis there, and composes mappings using multiple basis operators.

Reference

  • Wu et al., "Solving High-Dimensional PDEs with Latent Spectral Models," ICML 2023. Paper

Description

LSM uses cross-attention to encode input functions into a latent space, then constructs orthogonal basis functions in that latent space (inspired by classical spectral methods). The operator mapping is decomposed into multiple basis operators, enabling efficient learning for high-dimensional PDEs.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading the linked ICML 2023 paper and the issue description to understand the LSM architecture: cross-attention encoding, latent orthogonal basis construction, and multiple basis operators. Done means implementing the described LSM in NeuralOperators.jl and verifying that its latent spectral mappings work for high-dimensional PDE operator learning.

Written by the indexing model from the issue text.

Assessment

Tech stack
julia
Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.