A hybrid SSD-attention-MoE architecture, Cheems, is reported to outperform LLaMA3, Mamba2, and Jamba on matched-size language-model evaluations, with a proposed RoPE-for-SSD proof and a new dynamic-mask attention.
Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Wonderful Matrices: Combining for a More Efficient and Effective Foundation Model Architecture
A hybrid SSD-attention-MoE architecture, Cheems, is reported to outperform LLaMA3, Mamba2, and Jamba on matched-size language-model evaluations, with a proposed RoPE-for-SSD proof and a new dynamic-mask attention.