ChavyvAkvar/atma-10b-L40S-mbs16-atma-raven-titans__reg-baseline__distr-0__mem-1__win-0 Updated Jul 27 • 12
ATMA: Length-Invariant Language Modeling via Polar Attention and Gated-Delta Compression Memory Paper • 2606.25156 • Published Jun 23 • 2