OpenAI探讨基准得分与测试工具和设置的关系
OpenAI聊了基准测试的细节：得分高低不只靠模型，还跟你用的工具和设置有关，长期智能体还能靠压缩上下文学得更久。
OpenAI指出基准得分不仅反映模型本身，还依赖于测试工具和设置。对于长期运行智能体，保留推理和压缩上下文可帮助模型基于已有内容持续学习。该观点来自OpenAI官方博文，涉及评估方法论。
A benchmark score reflects the model as well as the harness and settings used to run it. For long-running agents, retaining reasoning and compacting context lets the model build on what it has already learned. openai.com/index/how-two-…📊 14 ⚡
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include OpenAI, 智能体, 基准评估, 上下文压缩.
Source attribution: curated from OpenAI (@OpenAI). Cross-source corroboration: 11 independent sources have covered this development (including arXiv: Anthropic, @OpenAIDevs, 宝玉), confirming industry-wide signal strength.
