TRINITY transferred zero-shot to four unseen tasks (AIME, BigCodeBench, MT-Bench, and GPQA). On average, the evolved coordinator surpassed every individual constituent model in its pool, including GPT-5, Gemini 2.5-Pro, and Claude-4-Sonnet.
作者声称一个仅20K参数的协调者能够超越GPT-5等顶级大模型，这一结论与行业对模型规模与能力关系的普遍认知相悖，提出了一个极具挑战性的反直觉观点。
The depth of recursion becomes a tunable compute axis at inference time, requiring no retraining. A small model, by reading itself, can iterate toward answers that neither it nor any of its workers could reach in a single pass.
大多数人认为模型的能力受其规模和训练数据的限制，需要更大模型或重新训练才能提升性能。但作者提出小模型通过自我递归调用可以在推理时动态扩展能力，无需重新训练就能达到单个模型无法企及的高度。这挑战了规模即能力的行业共识，暗示小模型可能通过自省机制实现突破性能力。
The depth of recursion becomes a tunable compute axis at inference time, requiring no retraining. A small model, by reading itself, can iterate toward answers that neither it nor any of its workers could reach in a single pass.
大多数人认为模型性能提升需要更大的参数规模或重新训练，但作者提出了一种反直觉的方法：通过递归调用自身，小模型可以在推理时自我迭代，达到单次推理无法达到的答案质量。这挑战了我们对模型规模与能力关系的传统认知。
On a 150-class benchmark, the surrogate fully replaces the teacher
大多数人认为复杂分类任务需要大型模型才能处理，小型代理模型只能处理简单任务。但作者展示了一个150类复杂任务中，小型代理模型完全能够替代教师模型，这挑战了'越大越好'的主流认知，证明了高效路由的潜力。
Because small, cheap, fast models are sufficient for much of the detection work, you don't need to judiciously deploy one expensive model and hope it looks in the right places. You can deploy cheap models broadly, scanning everything, and compensate for lower per-token intelligence with sheer coverage and lower cost-per-token.
这一观点提出了AI安全的经济新模式，通过广泛部署小型廉价模型来弥补单一大模型的不足。这种'广撒网'策略可能比依赖少数昂贵模型更有效，尤其在大规模代码库扫描场景中，为AI安全的经济可行性提供了新思路。
The capability rankings reshuffled completely across tasks. There is no stable best model across cybersecurity tasks. The capability frontier is jagged.
这一发现揭示了AI安全能力的'锯齿状前沿'现象，不同模型在不同安全任务上的表现差异巨大。这表明不存在'一刀切'的最佳安全模型，而是需要根据具体任务选择合适的模型，这对AI安全系统的设计有重要启示。
On a single H200 GPU with 1.5TB host memory, MegaTrain reliably trains models up to 120B parameters.
令人惊讶的是：仅使用一块配备1.5TB主机内存的H200 GPU就能训练1200亿参数的模型，这打破了人们对大规模模型必须依赖多GPU集群的固有印象。这一技术突破可能使超大规模模型训练变得更加普及和经济。
Without any architectural modification, MinerU2.5-Pro achieves 95.69 on OmniDocBench v1.6, improving over the same-architecture baseline by 2.71 points and surpassing all existing methods including models with over 200× more parameters.
大多数人认为更大的模型架构必然带来性能提升，但作者仅通过数据工程和训练策略优化，在保持1.2B参数架构不变的情况下，超越了参数量超过200倍的现有模型，这挑战了'越大越好'的行业共识，证明了数据质量的重要性。
Altmann, E. G. (2020). Spatial interactions in urban scaling laws. ArXiv:2006.14140 [Physics]. http://arxiv.org/abs/2006.14140