Noam Shazeer
VERIFIED TECHNICAL DOSSIERSources checked

Noam Shazeer

Distinguished AI Architect

2 Verified ArtifactsSource Checked & Attributed

Verified Proof of Work Artifacts

2 items cataloged

Each artifact below represents an authenticated research publication, production code repository, or technical architectural framework directly authored or co-created by Noam Shazeer. Every entry undergoes editorial source verification.

#1
RESEARCH Checked Sep 20, 2026

Fast Transformer Decoding: One Write-Head is All You Need (Multi-Query Attention)

Authored the foundational paper introducing Multi-Query Attention (MQA), where all attention heads share a single key and value head. This drastically slashes memory bandwidth consumption during autoregressive generation by up to 10x with negligible accuracy drop.

Model & Execution Context:Decoupled query/key/value projection matrices; foundation for Grouped-Query Attention (GQA) across modern LLMs like Llama 3 and Mistral.
Scope & Limitations

Slightly reduces representational capacity in small model regimes compared to standard multi-head attention before being mitigated by modern GQA configurations.

#2
RESEARCH Checked Sep 20, 2026

GLU Variants Improve Transformer (SwiGLU Activation Function)

Formulated and benchmarked Gated Linear Unit variants (GLU, Bilinear, ReGLU, GEGLU, SwiGLU) for transformer feedforward layers, demonstrating that SwiGLU consistently outperforms standard ReLU and GELU activations.

Model & Execution Context:Swish-gated linear projections: SwiGLU(x) = Swish(xW) * (xV).
Scope & Limitations

Requires three weight matrices in feedforward blocks instead of two, requiring adjustments to intermediate hidden dimensions to maintain equivalent parameter counts.