Author

A. Bocharnikov

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Apr 2026

KV Cache Offloading for Context-Intensive Tasks

This work evaluates modern KV offloading on Text2JSON and other context-intensive tasks and finds significant performance degradation on both Llama 3 and Qwen 3 models, highlighting the need for a comprehensive and rigorous evaluation of long-context compression techniques.

A. Bocharnikov, Ivan Ermakov, Denis Kuznedelev et al. · arXiv.org · 1 citation