Unverified50% confidenceFactExact time
2024年OpenAI o1的发布将Scaling Law扩展到推理阶段，允许模型生成更长内部思维链能带来与增大参数规模相近的性能提升
1
Sources
50%
Confidence
Long-term
Relevance
7/20/2026
First Seen
Sources
Related Claims
UnverifiedOpenAI o1 and DeepSeek-R1 achieve breakthroughs on complex benchmarks such as math competition problems by extending the chain of thought, which is essentially dynamically allocating more computational budget during the inference stage.78% similarUnverified推理强度机制的慢思考范式源自 OpenAI o1 系列，Anthropic 随后在 Claude 3.7 及更新模型中跟进实现77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/573358API
curl https://kongchang.com/api/v1/knowledge/claims/573358MCP
get_claim(id=573358)