Author

Steven Kolawole

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.

#machine learning Jun 2026

Epiphany-Aware KV Cache Eviction Without the Attention Matrix

EpiKV is introduced, which combines hidden-state shifts with the model's recent query--key relevance to rank cached tokens without materializing the attention matrix, and matches or outperforms the strongest attention-based eviction baselines while running directly in vLLM with unmodified attention kernels.

Steven Kolawole, Virginia Smith · arXiv.org · 2 citations