EchoKV：用迭代查询反馈自我修正 KV 缓存淘汰
EchoKV: Self-Corrective KV Cache Eviction via Iterative Query Feedback
arXiv 上一篇论文，讲 KV 缓存淘汰怎么用迭代反馈自我纠错，在 LongBench 上比 LAQ 强，做长上下文推理优化可以看看。
EchoKV 针对 KV cache 淘汰中的查询-工作区耦合问题提出迭代修正方法：先用未来解码位置的建议初始化 bootstrap 工作区，再用早期解码生成的查询修复工作区。它在 LongBench 的 16 个任务上测试，使用 Mistral-7B-Instruct-v0.2 和 Qwen2.5-7B-Instruct 两个骨干模型。在多跳跨文档 QA 和长单文档理解等证据分散的任务上，EchoKV 明显超过 LAQ。其前向传播次数与 LAQ 相同，只增加一次张量重打分操作。
EchoKV: Self-Corrective KV Cache Eviction via Iterative Query Feedback
KV cache eviction is not merely a token-ranking problem: the workspace retained for generating eviction queries also determines the quality of those queries. Consequently, an early workspace error can distort later query observations and reinforce an incorrect final eviction decision, particularly when evidence is dispersed across a long context. We present EchoKV, which discovers and addresses this failure mode with iterative correction. It initializes a feedback-compatible bootstrap workspace using proposals at future decode positions, then uses early decode-generated queries to recover the workspace before the remaining lookahead steps. It uses the same number of workspace forward passes as LAQ while adding a tensor re-scoring operation for workspace repair. On LongBench (16 tasks, Mistral-7B-Instruct-v0.2 and Qwen2.5-7B-Instruct), EchoKV shows notable improvements over LAQ on several tasks requiring dispersed evidence, including multi-hop cross-document QA and long single-document comprehension. Across budgets and backbones, EchoKV selectively improves evidence-intensive tasks while maintaining comparable performance on the remaining LongBench subsets. These results identify query--workspace coupling as a practical failure mode of query-driven KV eviction.