Unverified50% confidenceBenchmarkExact time
In a Kimi Linear, 8-card B200, concurrency-8 test, cached-prefix throughput improved by approximately 10.9x.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/3/2026
First Seen
Valid until: 1/1/2027
Sources
Related Entities
Related Claims
UnverifiedAblation analysis shows that fine-tuning on the segment paradigm yields a +5.29-point improvement on the test set, while replacing the first 4,500 transcript words with 2,000 retrieved words yields only a +1.55-point improvement.60% similarUnverifiedIn the Skatebench test, from the x-high tier to the max tier, reasoning tokens per question spiked from approximately 330 to nearly 5,000 (more than 10x), while scores improved by only 1%.58% similarUnverified运行10次测试置信区间宽度约±31%，100次收窄至±10%，400次进一步收窄至±5%，推荐关键测试用例至少运行20-50次57% similarUnverifiedMax级别相比extra high最多多消耗两倍的token，但在各类基准测试中仅带来4%至10%的性能提升57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/969381API
curl https://kongchang.com/api/v1/knowledge/claims/969381MCP
get_claim(id=969381)