An "American Exceptionalism" feature found in Meta's Llama-3.1-8B-Instruct. It controls the model's tendency to generate assertions of US superiority, a control absent in the Chinese model it was compared against.
令人惊讶的是，Anthropic 对美国模型同样一视同仁：在 Meta 的 Llama 中发现了「美国例外主义」特征。这说明政治偏向并非中国模型专属，而是所有大模型都可能内嵌的训练产物。研究团队以对称方式披露这两个发现，在政治上极为罕见，也极具勇气。
[Act 1996_108_178_20130201 amended by Act 2012_00A_010_20130201 wef 2013/08/13 ito GG20130822_36774]
Judicial Service Commission - composition - Constution s178 requires 23 members split between components with sep powers, being 10 from the legislature, 5 from the Executive, 3 from the Judiciary, and 5 from the Public. This split is a political one as it seems to follow no sensible basis.
Even our ability to detect online manipulation is affected by our political bias, though not symmetrically: Republican users are more likely to mistake bots promoting conservative ideas for humans, whereas Democrats are more likely to mistake conservative human users for bots.
Our brains work not that differently in terms of interconnectedness.Psychologists used to think of the brain as a limited storage spacethat slowly fills up and makes it more difficult to learn late in life. Butwe know today that the more connected information we alreadyhave, the easier it is to learn, because new information can dock tothat information. Yes, our ability to learn isolated facts is indeedlimited and probably decreases with age. But if facts are not kept
isolated nor learned in an isolated fashion, but hang together in a network of ideas, or “latticework of mental models” (Munger, 1994), it becomes easier to make sense of new information. That makes it easier not only to learn and remember, but also to retrieve the information later in the moment and context it is needed.
Our natural memories are limited in their capacities, but it becomes easier to remember facts when they've got an association to other things in our minds. The building of mental models makes it easier to acquire and remember new information. The down side is that it may make it harder to dramatically change those mental models and re-associate knowledge to them without additional amounts of work.
The mental work involved here may be one of the reasons for some cognitive biases and the reason why people are more apt to stay stuck in their mental ruts. An example would be not changing their minds about ideas of racism and inequality, both because it's easier to keep their pre-existing ideas and biases than to do the necessary work to change their minds. Similar things come into play with respect to tribalism and political party identifications as well.
This could be an interesting area to explore more deeply. Connect with George Lakoff.
Gugerty, L., Shreeves, M., & Dumessa, N. (2021). Biased belief updating in causal reasoning about COVID-19. PsyArXiv. https://doi.org/10.31234/osf.io/bfw76
Wang, X., Sirianni, A. D., Tang, S., Zheng, Z., & Fu, F. (2020). Public Discourse and Social Network Echo Chambers Driven by Socio-Cognitive Biases. Physical Review X, 10(4), 041042. https://doi.org/10.1103/PhysRevX.10.041042
Kasy, M. (2020). How to run an adaptive field experiment. Retrieved from https://maxkasy.github.io/home/files/slides/adaptive_field_slides_kasy.pdf
Mandalaywala, T. M., & Rhodes, M. (2020). Gender stereotypes about leadership in early childhood. https://doi.org/10.31234/osf.io/ytsbn
Ekstrom, P. D., & Lai, C. K. (2020, June 18). The Selective Communication of Political Information. https://doi.org/10.31234/osf.io/pnr9u