Unverified50% confidenceBenchmarkExact time
完整预训练后NCP-ArchPreview在下游任务宏观平均分上比OLMo-3-7B高出2.45分
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/11/2026
First Seen
Valid until: 12/10/2026
Sources
Related Entities
Related Claims
UnverifiedNCP-ArchPreview仅消耗51.3%的训练Token就达到了OLMo-3-7B的最终预训练损失75% similarUnverifiedNCP-ArchPreview在数学推理基准GSM8K上取得了5.99分的提升57% similarUnverified在SST-2数据集DistilBERT到DistilBERT配置中，可靠性门控方法达到88.800±0.532%，RAPID达到88.685±0.462%，相比基线87.271±0.162%提升约1.4个百分点57% similarUnverifiedEfficientNet 从 B0 到 B7，模型参数从 5.3M 增长到 66M，Top-1 精度从 77.1% 提升到 84.3%56% similarUnverifiedClaude在OSWorld基准测试上的任务完成率达到了14.9%，代表了当时的最高水平56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/907844API
curl https://kongchang.com/api/v1/knowledge/claims/907844MCP
get_claim(id=907844)