Unverified50% confidenceFactExact time
Unlimited OCR 2 uses the SR-SWA attention mechanism, keeping the decoding cost per step almost constant.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedUnlimited OCR 2's SR-SWA introduces a reference set mechanism on top of the sliding window, enabling text and visual information to interact across windows and avoiding the problem of pure sliding windows forgetting distant information.71% similarUnverifiedAfter digging into the code, community members pointed out that SWA Pounday Replay is a technique for efficiently managing attention windows in long-context inference that can greatly reduce the memory footprint of the KV cache.65% similarUnverified在性能敏感的代码中，显式使用view并手动管理连续性优于使用reshape，因为reshape可能带来额外的内存开销和计算时间60% similarUnverifiedPrefix caching, by reusing KV vector caches for identical context prefixes, can bring 50%–80% computational savings in AI agent scenarios and significantly reduce time-to-first-token (TTFT) latency.59% similarUnverifiedoMLX employs Continuous Batching technology, re-evaluating batch composition at every decoding step.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102916API
curl https://kongchang.com/api/v1/knowledge/claims/1102916MCP
get_claim(id=1102916)