Unverified70% confidenceBenchmarkExact time
Chinchilla surpassed Gopher, which has 280 billion parameters, on multiple benchmarks with only 70 billion parameters.
2
Sources
70%
Confidence
Long-term
Relevance
7/25/2026
First Seen
Sources
Related Claims
UnverifiedChinchilla模型以70B参数配合4倍数据量即可超越280B参数的Gopher模型性能78% similarUnverifiedThe 70-billion-parameter Chinchilla model, trained on 1.4 trillion tokens, outperformed Gopher, which has 280 billion parameters but was trained on only 300 billion tokens.71% similarUnverifiedThe Chinchilla model, with 70B parameters and 1.4T tokens, significantly outperforms the Gopher model, which has 280B parameters and 300B tokens, under the same compute budget.69% similarUnverifiedThe model has a core of 125B parameters, plus an n-gram lookup table with 51B parameters and a multi-token prediction (MTP) draft layer with 4B parameters, totaling approximately 180B.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/615293API
curl https://kongchang.com/api/v1/knowledge/claims/615293MCP
get_claim(id=615293)