Unverified50% confidenceFactExact time
NLP领域的GLUE（2018）和SuperGLUE（2019）等综合性基准强化了以标准化基准比较模型的趋势
1
Sources
50%
Confidence
Long-term
Relevance
7/11/2026
First Seen
Sources
The Pain of AI Paper Peer Review: Why Should Models Released After the Deadline Count Against You?
redditr/LanguageTechnology7/9/2026
Related Claims
UnverifiedNLP领域的GLUE于2018年、SuperGLUE于2019年推出，作为综合性基准强化了刷榜趋势84% similarUnverified智谱GLM系列模型自2022年起持续迭代，与Meta的LLaMA、Mistral等西方开源模型竞争66% similarUnverified2022-2023 年间「大模型+工具调用」的可行性被相继验证65% similarUnverifiedGPTQ was proposed in 2022 and improves INT4 quantization quality through layer-wise second-order information calibration.64% similarUnverifiedAs of 2025, domestically developed large models have reached internationally top-tier levels across multiple benchmark tests.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/484316API
curl https://kongchang.com/api/v1/knowledge/claims/484316MCP
get_claim(id=484316)