GPT-6.1 Sol 减少事实错误响应
OpenAI 新版 GPT-6.1 Sol 事实错误减少 32%，安全审查未被绕过，比前代更可靠。
OpenAI 发布 GPT-6.1 Sol，在故意设计的事实性提示测试中，相比 GPT-6 Sol，在低推理努力下减少了约 32% 的错误响应。
在自动化安全审查评估中，未观察到尝试绕过安全审查的行为。该模型在事实准确性方面有明显改进。
On deliberately difficult factuality prompts, GPT-6.1 Sol reduces responses containing factual errors by about 32% versus GPT-6 Sol at low reasoning effort. In our automated safety review evaluation, we observed no attempts to bypass the safety reviewer.📊 3 ⚡
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include GPT-6.1 Sol, OpenAI, factual accuracy, reasoning models.
Primary source verification: published directly by @OpenAIDevs with first-party authority. Cross-source corroboration: 11 independent sources have covered this development (including OpenAI, rohanpaul_ai, Techsauce), confirming industry-wide signal strength.
