Unverified50% confidenceFactExact time
v0.30.0 introduces HiSparse, which adds a host tier beneath sparse MLA decoding to form a hierarchical cache structure.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/24/2026
First Seen
Valid until: 12/23/2026
Sources
Related Entities
Related Claims
UnverifiedvLLM v0.28.0rc2 introduces a new speculative decoding feature called DFlash2, which combines two major technical components: local convolution and a candidate selector (PR #52816).60% similarUnverifiedCID有v0和v1两个版本，v0使用base58编码（以Qm开头），v1使用multibase编码支持更多哈希函数和编解码器60% similarUnverifiedv0.29.0rc6 contains one change, "Apply dense prefix cache default to hybrid models," which sets dense prefix caching as the default strategy for hybrid models.56% similarUnverifiedv0.30.0 simplifies the execution flow related to KDA, AttnRes, and MLA, and adds MXFP8 KV cache and asynchronous Engram mechanisms for some models.56% similarUnverifiedCgroups v2 was introduced in Linux 4.5, adopting a unified hierarchy structure to replace the multi-hierarchy design of v1.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/945472API
curl https://kongchang.com/api/v1/knowledge/claims/945472MCP
get_claim(id=945472)