Unverified50% confidenceBenchmarkExact time
11 of the 18 models correctly answered at least 96% of memory read questions, but the gap widened significantly on the memory write task (maintaining the passport).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/11/2026
First Seen
Valid until: 1/9/2027
Sources
Related Entities
Related Claims
UnverifiedIn memory recall tests, the 12B model answered 82 out of 100 questions correctly and the 31B model answered 98 out of 100, with neither model hallucinating answers.70% similarUnverifiedIn real agent workloads, approximately 96% of requests are cache reads.70% similarUnverified端侧推理中吞吐量需达到5-10 token/s对应人类舒适阅读速度，低于此值输出显得一字一顿67% similarUnverifiedDeliberately implanted false memories in AI agents spread to other users at a probability of 6% to 38%.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/997246API
curl https://kongchang.com/api/v1/knowledge/claims/997246MCP
get_claim(id=997246)