Unverified50% confidenceOpinionExact time
讨论者的共识是如果基座模型在pass@16上的成功率为0，RL几乎不可能凭空创造出新能力，但只要存在非零成功概率RL就能将其增强放大
1
Sources
50%
Confidence
Long-term
Relevance
8/13/2026
First Seen
Sources
Related Claims
Unverified研究表明，即使能力相对较弱的模型，其pass@10或pass@20的表现往往能够接近强模型的pass@1水平68% similarUnverifiedOpenAI's chart shows that at lower test-time compute levels, the model's success rate on this problem is close to zero, with occasional success only under extremely long reasoning chains.63% similarUnverifiedThe generational capability gap between models is narrowing, with multiple models approaching or even surpassing GPT-4's reasoning advantage on certain benchmarks.61% similarUnverifiedThe average evaluation metrics of a new LLM may be better than those of an older model, but regressions can occur in niche critical scenarios.61% similarUnverifiedFor small models with limited parameter counts, data quality can have an even greater impact on final performance than data quantity.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/742561API
curl https://kongchang.com/api/v1/knowledge/claims/742561MCP
get_claim(id=742561)