Treffer: Longer Attention Span: Increasing Transformer Context Length With Sparse Graph Processing Techniques

Title:
Longer Attention Span: Increasing Transformer Context Length With Sparse Graph Processing Techniques
Authors:
Tomczak, Nathaniel, Kuppannagari, Sanmukh
Source:
2025 IEEE International Parallel and Distributed Processing Symposium (IPDPS) IPDPS Parallel and Distributed Processing Symposium (IPDPS), 2025 IEEE International. :1-12 Jun, 2025
Relation:
2025 IEEE International Parallel and Distributed Processing Symposium (IPDPS)
Database:
IEEE Xplore Digital Library