implementation detail · filed under model architecture
Per-layer-type rotary position scales
Rotary position scales that vary by layer type.
Also called per-layer-type rotary scales.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- core
How sources treat it
One count per evidence span, weakest treatment to strongest.
core 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
with softplus attention gating and per-layer-type rotary scales
coremodel architecturein Laguna S 2.1Poolside
Filed alongside
Other methods under model architecture :: positional encoding.
YaRNRotary Position Embedding2D rotary position embeddingRoPE scalingLearned input-dependent relative position biasNo Position EncodingProportional Rotary Position EmbeddingRelative attention2D coordinate-based positional embeddingsGated attention with partial RoPEOmitting RoPE in attention layersPartial RoPETemporal-Modality Rotary Position Embedding