AvoKV-E: Payload-Aware KV Cache Eviction for Long Reasoning

arXiv:2610.03007v1 Announce Type: cross Abstract: Long-output reasoning shifts the KV-cache bottleneck from the fixed prompt to the generated trace. Existing reasoning-cache eviction methods largely treat cached entries as routing objects, estimating whether an old key will still be read, will recur, or can be replaced. This routing-only view overlooks two effects: low-attention entries can carry large value payloads whose removal changes future predictions, and newly generated states can appear stale before later queries have had a chance to read them. We introduce AvoKV-E, a training-free eviction policy that first delays eligibility for recent states and then ranks eligible entries using candidate-normalized read pressure, key redundancy, and value-payload potential. According to empirical evaluation across different models and datasets, AvoKV-E matches or exceeds redundancy-aware, recurrence-based, and thought-adaptive eviction baselines at matched active-KV budgets, with its largest gains in the tightest-cache regime. Component and counterfactual analyses further connect these gains to delayed observation, payload-aware scoring, redundancy, and scale-robust normalization. Together, the results show that long-reasoning KV eviction should preserve not only keys that are likely to be read, but also the value payloads that sustain the reasoning trajectory.

Related

• Fractional Decay KV-Cache: Ownership-Aware Memory Management for Improved Inference Relevancy in Dialog Systems
• Adaptive Filtering of the KV Cache: Diagnosing and Correcting Structural-Role Bias in LLM Inference
• Kara: Efficient Reasoning LLM Serving via Sliding-Window KV Cache Compression

Source: arXiv cs.AI | 2026-10-05