3 papers
cs.IT2026
Polynomial Context-Truncation Sensitivity in Autoregressive Language Models: Sequential Wyner-Ziv Bounds for KV Cache Compression
Munsik Kim
We study the rate-distortion limits of online KV cache compression in autoregressive language models, formulating it as sequential Wyner-Ziv source coding on the filtration induced…
math.ST2026
A Temporal Spatial Minimax Rate for Smoothly-Varying Distributions in Wasserstein Space
Munsik Kim
We study the minimax rate of estimating a future value of a curve in the -Wasserstein space from finitely many noisy sn…
cs.IT2026
Information-Theoretic Lower Bounds for Bit-Constrained Stochastic Optimization via a Reduction to Compressed Gaussian Mean Estimation
Munsik Kim
Low-precision pretraining (FP8, MXFP4, NVFP4) is now standard for frontier language models, yet the literature is almost entirely achievability -- algorithms and empirical scaling…