SIREN-RoPE learns the rotation space in RoPE from timestamps, cyclical patterns, and categorical signals using SIREN, improving calibration and ranking in a production news recommender with negligible overhead.
All bases produce a monotone global decay with small high-frequency oscillations whose period grows with the base
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning to Rotate: Temporal and Semantic Rotary Encoding for Sequential Modeling
SIREN-RoPE learns the rotation space in RoPE from timestamps, cyclical patterns, and categorical signals using SIREN, improving calibration and ranking in a production news recommender with negligible overhead.