Unverified50% confidenceBenchmarkExact time
The report states that MNN-Pocket-0.3B reaches the frontier level among models of the same size and surpasses some 0.5B-parameter models on instruction-following benchmarks.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
11/10/2026
First Seen
Valid until: 11/24/2026
Sources
Related Entities
Related Claims
Unverified研究表明当模型参数规模达到3B以上时,1.58-bit模型在多项基准测试中表现可接近甚至匹配同等规模的全精度模型72% similarUnverified万象模型提供了 1.3B 和 14B 参数量的不同规模版本68% similarUnverifiedThe receiver small model used in the CacheGraft experiments has 0.6B parameters, and the source large model has 4B parameters.66% similarUnverifiedIn practice, 4-bit to 8-bit quantized versions perform nearly identically to the original model on most tasks.65% similarUnverified在27B混合架构、4B稠密模型、36B MoE三类模型上以8k上下文测试,Prefill阶段性能无变化,Decode阶段打补丁后快约1.4%65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1154771API
curl https://kongchang.com/api/v1/knowledge/claims/1154771MCP
get_claim(id=1154771)