← Search

Pengkun Jiao

2 accepted papers

2026

Predicting Future KV Utility: Global Combinatorial Optimization for Task-Agnostic KV Cache Eviction

ICML 2026poster

Given the quadratic complexity of attention, KV cache eviction is vital to accelerate model inference. Current KV cache eviction methods typically rely on instantaneous heuristic metrics, implicitly assuming that score magnitudes are consistent proxies for importance across all heads. However, this …

Cited by 0SourceScholar
2025

From Holistic to Localized: Local Enhanced Adapters for Efficient Visual Instruction Fine-Tuning

ICCV 2025poster

Efficient Visual Instruction Fine-Tuning (EVIT) seeks to adapt Multimodal Large Language Models (MLLMs) to downstream tasks with minimal computational overhead. However, as task diversity and complexity increase, EVIT faces significant challenges in resolving data conflicts. To address this limitati…