Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
ResidualQuant: KV Cache Quantization for Looped Transformers with 2-Bit Residuals
Heejun Kim, Junyoung Lee, SangLyul Cho +3
Looped Transformers improve parameter efficiency by applying shared Transformer blocks over recurrent loops, increasing computational depth without increasing the parameter count.…
cs.LG2026
LoopSpec: Pipelined Self-Speculative Decoding for Looped Transformers
SangLyul Cho, Langqing Cui, Sehoon Kim +2
Looped Transformers achieve strong performance with compact parameter sizes by repeatedly applying a shared stack of Transformer blocks across recurrent depths. However, they incur…