# ASPIRE: Make Spectral Graph Collaborative Filtering Great Again via Adaptive Filter Learning

arXiv: https://arxiv.org/abs/2604.22549 · HTML: https://mlsanity.com/paper/2604.22549/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Apr 2026.

## Built on by (most cited first, 1 of 1)

- [Rethinking Semantic Alignment in LLM-Enhanced Collaborative Filtering: A Spectral Decoupling Approach](https://mlsanity.com/paper/2608.24363/) (2026)

## Abstract (arXiv)

Graph filter design is central to spectral collaborative filtering, yet most existing methods rely on manually tuned hyperparameters rather than fully learnable filters. We show that this challenge stems from a bias in traditional recommendation objectives, which induces a spectral phenomenon termed low-frequency explosion, thereby fundamentally hindering the effective learning of graph filters. To overcome this limitation, we propose a novel adaptive spectral graph collaborative filtering framework (ASPIRE) based on a bi-level optimization objective. Guided by our theoretical analysis, we disentangle the filter learning objective, which in turn leads to excellent recommendation performance, spectral adaptivity, and training stability in practice. Extensive experiments show our learned filters match the performance of carefully engineered task-specific designs. Furthermore, ASPIRE is equally effective in LLM-powered collaborative filtering. Our findings demonstrate that graph filter learning is viable and generalizable, paving the way for more expressive graph neural networks in collaborative filtering.