{"id":"arxiv:2604.25975","kind":"preprint","title":"Rethinking KV Cache Eviction via a Unified Information-Theoretic Objective","authors":["Jiaming Yang","Chenwei Tang","Liangli Zhen","Jiancheng Lv"],"author_count":4,"year":2026,"published":"2026-04-28","venue":null,"cited_by":null,"doi":null,"arxiv_id":"2604.25975","nct_id":null,"source_url":"https://arxiv.org/abs/2604.25975","archive_url":"https://space-kitty.com/papers/arxiv:2604.25975","updated":"2026-08-10","pmid":null,"pmcid":null,"openalex_id":null,"pdf_url":"https://arxiv.org/pdf/2604.25975","abstract":"Key-Value (KV) caching is essential for large language model inference, yet its memory overhead poses a critical bottleneck for long-context generation. Existing eviction policies predominantly rely on empirical heuristics, lacking a rigorous theoretical foundation. This work rethinks KV cache eviction through the lens of the Information Bottleneck principle. Under a linear-Gaussian surrogate of attention, we derive a closed-form mutual information objective that characterizes the effective information capacity of a retained KV cache subset. This formulation reveals that a wide range of existing eviction strategies can be interpreted as different approximations of the same capacity-maximization principle. Guided by this insight, we introduce CapKV, a capacity-aware eviction method that directly targets information preservation via a log-determinant approximation using statistical leverage scores. This approach replaces heuristic selection with a theoretically grounded mechanism that preserves the maximum predictive signal. Extensive experiments across multiple models and long-context benchmarks show that CapKV consistently outperforms prior methods, achieving a better trade-off between memory efficiency and generational fidelity. Our code is available at https://github.com/jiamingyy/CapKV","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":false,"retracted":false,"peer_reviewed":false,"categories":["cs.LG","cs.AI","cs.IT"],"details":null,"provenance":"arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.","topics":[],"cited_in_dossiers":[]}