Unverified50% confidenceFactExact time
《Attention Is All You Need》论文在发表初期并未全面碾压所有基线，却奠定了Transformer时代的架构基础
1
Sources
50%
Confidence
Long-term
Relevance
7/11/2026
First Seen
Sources
AI论文同行评审之痛：截止日期后的模型凭什么成扣分理由？
redditr/LanguageTechnology7/9/2026
Related Claims
VerifiedTransformer最初由Vaswani等人在2017年的论文《Attention Is All You Need》中提出，用于机器翻译任务69% similarUnverifiedLlion Jones是开创性论文《Attention Is All You Need》的八位作者之一，该论文提出了Transformer架构68% similarPartially VerifiedLlion Jones是Transformer论文'Attention Is All You Need'的共同作者之一65% similarUnverifiedNoam Shazeer是Transformer架构论文'Attention Is All You Need'的共同作者之一，该论文于2017年发表63% similarUnverified线性注意力思路由 Katharopoulos 等人在 2020 年论文《Transformers are RNNs》中系统化63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/484098API
curl https://kongchang.com/api/v1/knowledge/claims/484098MCP
get_claim(id=484098)