Unverified50% confidenceOpinionExact time
模型的基准分数更高并不等于在所有场景下都更好用
1
Sources
50%
Confidence
Long-term
Relevance
7/7/2026
First Seen
Sources
Related Claims
Unverified对比学习中的推远操作只要求负样本相似度低于正样本，并不要求达到-1的极端值68% similarUnverifiedFor small models with limited parameter counts, data quality can have an even greater impact on final performance than data quantity.68% similarUnverifiedWhen benchmark scoring metrics become optimization targets, a Goodhart's Law effect arises: models learn to score high on specific formats without necessarily truly improving their capabilities, and some models with HumanEval scores exceeding 95% vary dramatically in actual code generation ability.67% similarUnverified评估模型不能只看智能指标，还要看Token使用效率和工具调用次数，工具调用次数过多会导致上下文污染67% similarUnverified综合多个独立来源的评测结果比单一自报基准分数更能反映模型的实际水平66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/126888API
curl https://kongchang.com/api/v1/knowledge/claims/126888MCP
get_claim(id=126888)