MSA eric8810 · COMPLETED
Memory Sparse Attention - A scalable, end-to-end trainable latent-memory framework for 100M-token contexts.
github.com/eric8810/MSA · ★ 0 · Forks 0 · Size 1.3 MB
SUMMARY
Technologies 5
Scored 5
Observed 5
Practices 6
Evidence 6
Skips 0
COVERAGE
Analyzed 40 files · 68 commits · 0 API calls
TECHNOLOGIES & DEPTH
Markdown LANGUAGE Depth 70
3 files · PRODUCTION
Shell LANGUAGE Depth 70
3 files · PRODUCTION
Python LANGUAGE Depth 70
26 files · PRODUCTION
pip BUILD_TOOL Depth 80
1 files · CONFIGURATION
NumPy LIBRARY Depth 66
3 files · PRODUCTION
PRACTICES
documentation · observedautomated_tests · absentcontinuous_integration · absentcontainerization · absentlinting · absentformatting · absent
ACTIVITY & OWNERSHIP
First commit 2025-10-29
Last commit 2026-04-05
Active months 4
Commits 68