a harness encodes an assumption about what the model can't do on its own
这一洞见是Agent工程演进的底层逻辑:脚手架是对模型当前能力边界的妥协。随着基座模型能力跃升,曾经的“必要组件”可能沦为冗余开销。因此,解构并剔除过时假设,是保持系统简洁高效的关键。
a harness encodes an assumption about what the model can't do on its own
这一洞见是Agent工程演进的底层逻辑:脚手架是对模型当前能力边界的妥协。随着基座模型能力跃升,曾经的“必要组件”可能沦为冗余开销。因此,解构并剔除过时假设,是保持系统简洁高效的关键。
Creating a community network ontology is therefore about much more than just knowledge representation. It also requires us to think about how this conceptual knowledge model affects real-world knowledge creation and application processes, in our case concerning participatory community network mapping. Its participatory nature means that we need to think hard about how to explicitly involve the community in the construction, evolution, and use of the ontology.
Smaldino, Paul E., and Cailin O’Connor. ‘Interdisciplinarity Can Aid the Spread of Better Methods Between Scientific Communities’. MetaArXiv, 5 November 2020. https://doi.org/10.31222/osf.io/cm5v3.
Candido, D. S., Claro, I. M., Jesus, J. G. de, Souza, W. M., Moreira, F. R. R., Dellicour, S., Mellan, T. A., Plessis, L. du, Pereira, R. H. M., Sales, F. C. S., Manuli, E. R., Thézé, J., Almeida, L., Menezes, M. T., Voloch, C. M., Fumagalli, M. J., Coletti, T. M., Silva, C. A. M. da, Ramundo, M. S., … Faria, N. R. (2020). Evolution and epidemic spread of SARS-CoV-2 in Brazil. Science. https://doi.org/10.1126/science.abd2161