Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
Why SWAVE May Not Be All You Need:A Concept-Evolution Retrospective on Complex-Valued Recurrent Language Models
Ramprasath Ganesaraja, Swathika N, Sahil Dilip Panse
SWave is a complex-valued recurrent language model (169.26M parameters, D=384, L=16, T=2048) trained on FineWeb-Edu using 2xH100 NVL. It was designed around three founding premises…
cs.LG2026
Ternary Mamba: Grouped Quantization-Aware Training of W1.58A16 State Space Models
Ramprasath Ganesaraja, Sahil Dilip Panse, Swathika N
State Space Models (SSMs) such as Mamba-2 offer linear-time inference but their memory footprint limits edge deployment. Prior ternary SSM work (Slender-Mamba) trains from scratch…