Unverified50% confidenceBenchmarkExact time
在HumanEval、MBPP、GSM8K等基准测试上，Gemma 7B的表现显著优于同参数量级的LLaMA 2 7B和Mistral 7B
1
Sources
50%
Confidence
Long-term
Relevance
7/16/2026
First Seen
Sources
Related Claims
UnverifiedMistral 7B证明精心训练的7B模型可以匹敌早期的13B甚至30B模型75% similarUnverifiedMistral AI的7B模型在发布时性能超越了两倍参数量的Llama 2 13B71% similarUnverifiedMistral的7B模型在发布时超越了Llama 2 13B的表现69% similarUnverifiedMistral AI's Mixtral 8x7B adopts a MoE architecture with a total parameter count of approximately 47B, but only approximately 13B parameters are actually activated during each inference.67% similarUnverifiedThe MiniCPM series achieves performance close to 7B and even 13B models on multiple benchmarks with a parameter count at the 2B–4B level66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/530558API
curl https://kongchang.com/api/v1/knowledge/claims/530558MCP
get_claim(id=530558)