12 Matching Annotations
  1. Feb 2019
    1. Maximum Entropy Generators for Energy-Based Models

      【能量视角下的GAN模型】本文直接受启发于Bengio团队的新作《Maximum Entropy Generators for Energy-Based Models》,作者给出了GAN/WGAN的清晰直观的能量图像,讨论了判别器(能量函数)的训练情况和策略,指出了梯度惩罚一个非常漂亮而直观的能量解释。此外,本文还讨论了GAN中优化器的选择问题。http://t.cn/EcBIwqJ

  2. Dec 2018
    1. A Probe into Understanding GAN and VAE models

      paper 提出了个 VAE-GAN 模型,不过正如作者自己说的可能是 GPU 资源不够,图像质量并不太如意,而且用的是 FCN 不是 CNN;主要用 Entropy 来量化评估生成变现。

    1. In essence, for a change to occur, you must apply more energy to the system than is extracted by the system.
  3. Oct 2018
    1. entropic

      This is what Edgar Orrin Klapp meant when he wrote in his 1986 Overload and Boredom: Essays on the Quality of Life in the Information Society that “meaning and interest are found mostly in the mid-range between extremes of redundancy and variety-these extremes being called, respectively, banality and noise” (). Redundancy is repetition of the same, which creates a condition of insufficient difference, while noise is the chaos of non-referentiality, or entropy. In a way, these extremes collapse into each other, in that both can be viewed “as a loss of potential … for a certain line of action at least” ().

      There is perhaps something of "the real" here, as well. Volker Woltersdorff (2012, 134) writes that: The law of increasing entropy is a concept of energy in the natural sciences that assumes the tendency of all systems to eventually reach their lowest level of energy. Organic systems therefore tend toward inertia … Freud identifies the death drive with entropy … within his theory, the economy of the death drive is to release tension."

      Adam Phillips clarifies the death drive: “People are not, Freud seems to be saying, the saboteurs of their own lives, acting against their own best interests; they are simply dying in their own fashion (to describe someone as self-destructive is to assume a knowledge of what is good for them, an omniscient knowledge of the ‘real’ logic of their lives)” (2000, 81, cf. 77).

  4. Sep 2018
    1. The concept of “extropy” was used to encapsulate the core values and goals of transhumanism. Intended not as a technical term opposed to entropy but instead as a metaphor, extropy was defined as “the extent of a living or organizational systems intelligence, functional order, vitality, and capacity and drive for improvement.”

      It's interesting that the author chooses to emphasize the distinction between extropy as an opposition to entropy, but instead as a metaphor. However, would extropy not be the opposite of entropy metaphorically as well? Scientific definition aside, entropy is defined as the universe's tendency towards chaos in all manners of the word meaning constant expansion and disorder. Extropy is the universe's tendency towards the idea of a 'singularity' (discussed in "The Technological Singularity") so essentially the exact opposite? The universe's tendency to follow "intelligence, functional order", etc. toward a single point of "posthuman" where we've gone beyond human capability?

  5. Jun 2018
    1. entropic

      This is what Edgar Orrin Klapp meant when he wrote in his 1986 Overload and Boredom: Essays on the Quality of Life in the Information Society that “meaning and interest are found mostly in the mid-range between extremes of redundancy and variety-these extremes being called, respectively, banality and noise” (). Redundancy is repetition of the same, which creates a condition of insufficient difference, while noise is the chaos of non-referentiality, or entropy. In a way, these extremes collapse into each other, in that both can be viewed “as a loss of potential … for a certain line of action at least” ().

      There is perhaps something of "the real" here, as well. Volker Woltersdorff (2012, 134) writes that: The law of increasing entropy is a concept of energy in the natural sciences that assumes the tendency of all systems to eventually reach their lowest level of energy. Organic systems therefore tend toward inertia … Freud identifies the death drive with entropy ... within his theory, the economy of the death drive is to release tension."

      Adam Phillips clarifies the death drive: “People are not, Freud seems to be saying, the saboteurs of their own lives, acting against their own best interests; they are simply dying in their own fashion (to describe someone as self-destructive is to assume a knowledge of what is good for them, an omniscient knowledge of the ‘real’ logic of their lives)” (2000, 81, cf. 77).

  6. Feb 2018
    1. Se constituye de galaxias, de astros, de soles, dicho de otro modo, se desarrolla mediante la organización al mismo tiempo que se produce mediante la desorganización. El mundo biológico es un mundo que evoluciona

      De hecho, algunos estudios sugieren que la organización es una buena manera de acelerar la desorganización, particularmente en el caso de la vida (https://www.quantamagazine.org/a-new-thermodynamics-theory-of-the-origin-of-life-20140122/)

  7. Dec 2017
    1. Life conforms to neither of these conditions. Take self-reliance. Living systems exist far from the state known as thermodynamic equilibrium – instead of their energy spreading itself out over the widest possible space, it’s concentrated in specific areas and flows along defined pathways, such as the cardiovascular or nervous system. Such phenomena are very improbable, as far as fundamental physics is concerned. Maintaining this unusual arrangement requires constant activity, or metabolism, which in turn demands that organisms extract energy from their environment via eating, breathing, photosynthesis, and so on. That belies any pretensions of independence.

      May be this the reason why life "goes against" entropy?

  8. Jan 2014
    1. An effective data management program would enable a user 20 years or longer in the future to discover , access , understand, and use particular data [ 3 ]. This primer summarizes the elements of a data management program that would satisfy this 20-year rule and are necessary to prevent data entropy .

      Who cares most about the 20-year rule? This is an ideal that appeals to some, but in practice even the most zealous adherents can't picture what this looks like in some concrete way-- except in the most traditional ways: physical paper journals in libraries are tangible examples of the 20-year rule.

      Until we have a digital equivalent for data I don't blame people looking for tenure or jobs for not caring about this ideal if we can't provide a clear picture of how to achieve this widely at an institutional level. For digital materials I think the picture people have in their minds is of tape backup. Maybe this is generational? New generations not exposed widely to cassette tapes, DVDs, and other physical media that "old people" remember, only then will it be possible to have a new ideal that people can see in their minds-eye.

    2. A key component of data management is the comprehensive description of the data and contextual information that future researchers need to understand and use the data. This description is particularly important because the natural tendency is for the information content of a data set or database to undergo entropy over time (i.e. data entropy ), ultimately becoming meaningless to scientists and others [ 2 ].

      I agree with the key component mentioned here, but I feel the term data entropy is an unhelpful crutch.

    3. data entropy Normal degradation in information content associated with data and metadata over time (paraphrased from [ 2 ]).

      I'm not sure what this really means and I don't think data entropy is a helpful term. Poor practices certainly lead to disorganized collections of data, but I think this notion comes from a time when people were very concerned about degradation of physical media on which data is stored. That is, of course, still a concern, but I think the term data entropy really lends itself as an excuse for people who don't use good practices to manage data and is a cover for the real problem which is a kind of data illiteracy in much the same way we also face computational illiteracy widely in the sciences. Managing data really is hard, but let's not mask it with fanciful notions like data entropy.