Rock Matrix 开源 EdgeRock-4B：固定内存的图文推理小模型
Rock Matrix 开源了个 4B 小模型 EdgeRock-4B，KV cache 大小固定，长推理不吃显存，AIME 上比同类固定状态模型高不少，做端侧部署的可以看看。
Rock Matrix 开源了 4B 参数的图文推理模型 EdgeRock-4B。该模型将注意力 KV cache 固定在 1,024 个 token 的窗口内，更早的重要信息存入独立记忆库，避免长对话和长推理时显存持续增长。官方报告其在 AIME 2024–2026 上的平均 pass@1 达 82.9%，而它测试过的最强固定状态模型仅为 18.8%。
Rock Matrix has open-sourced EdgeRock-4B, a small text-and-image reasoning model with fixed-capacity memory.
EdgeRock-4B keeps its attention KV cache fixed in size by holding the last 1,024 tokens in full and saving important older details to a separate bank.
Standard language models keep a growing record of every token they read, so long chats and long reasoning take up more GPU memory over time.
Rock Matrix reports an 82.9% mean pass@1 on AIME 2024–2026, compared to 18.8% for the strongest bounded-state model it tested.
