a symbolic-logic-based Feasibility Memory utilizes executable Python verification functions synthesized from failed transitions
大多数人认为LLM应该从成功经验中学习,但作者提出从失败过渡中合成验证函数的观点极具反直觉。这种方法将失败视为宝贵资源而非需要避免的问题,挑战了机器学习领域的主流优化思想。
a symbolic-logic-based Feasibility Memory utilizes executable Python verification functions synthesized from failed transitions
大多数人认为LLM应该从成功经验中学习,但作者提出从失败过渡中合成验证函数的观点极具反直觉。这种方法将失败视为宝贵资源而非需要避免的问题,挑战了机器学习领域的主流优化思想。
The worry most people have with this suggestion is that children are going to get discouraged if they fail. But that is not necessarily the case, and I think teachers, parents, and other adults have a great opportunity to help prevent this. If we demonstrate that needing to put down a book for awhile is not a failure, then we can help children become more willing to experiment and to try things which are currently just out of reach.
This is the concept of growth mindset; and we need to teach that to our children in any way possible. It has been shown in studies that growth mindset has a positive causal influence on academic and financial success (I cannot state sources, but I know I've come across this)
Note to self: Research this later.
Children need to learn to read difficult books, or else once they are in college they won’t be able to do so. That probably means that they need to attempt to read some of these books, even when we know they will likely fail.
Success is the ability to go from failure to failure without losing enthusiasm (or motivation)
The sixth step, most essential as well, is to Accept the Wins
Owning the losses means also owning the wins.
The fifth step is to have Selective Memory only choose to remember the events that serve the future. Things that help to improve in the future.
It's like Marcus Aurelius wrote (in a slightly different way): "Ask yourself at any moment, is this essential?" In this way it would become: "Ask yourself at any moment, does this help me?"
The fourth step is to Apply the Reflection. Adjust behavior based on reflection. We improve not for validation, we improve for ourselves (stoic philosophy)
Document the journey in for example a journal. Make a comparison between what would be done in the past and what will be done in the future.
Data collection. Measurement.
Marginal Gains. It's sort of a daily continous Kolb's cycle but in a more lightweight form. I can already see the power in this. Absolute gem.
Could also be overwhelming if applied to a lot. therefore, use the power law and focus on what is essential to life change. (thanks Dr. Benjamin Hardy.)
The third step is to Reflect and think into the future. Extract meaning and lessons from the failure. Think about opportunities.
Reflection increases confidence. Kolb's can help with this a lot.
The second step is Sit with the loss in order to find the (root) cause of the loss or pain. Do not avoid the pain, don't distract oneself, instead embrace it and feel it.
Endurance can be trained. Comfort with uncomfortability can be trained in the same way.
Accept and sit in the fire. Embrace the turmoil.
The first step to deal with loss of any kind, be it a girlfriend, love, job, purpose, etc. Is to ACCEPT YOU LOST
Failure = Failure.
Failure is inevitable, and will be part of any learning process. Therefore it should not be avoided at all costs. It should be used to learn from. However; there is also no point in seeking failure, for if failure is not something negative, there is no point to improve (says the author at least)
Pescetelli, N., Rutherford, A., Kao, A., & Rahwan, I. (2020). Pescetelli_et_al_Collective learning in news consumption. https://doi.org/10.31234/osf.io/w6nc5
In Medias Res teams: The case for crisis learning alongside government covid-19 responses. (2020, April 13). Observatory of Public Sector Innovation. https://oecd-opsi.org/inmediasres/
“The chief trick to making good mistakes is not to hide them — especially not from yourself.”
Quote from Daniel Dennett