• Is Causal AI simply ReLU (Geoffrey E. Hinton 2010) (Re: Mark V. Shaney caused Julio Di Egidio Suicide)

    From Mild Shock@janburse@fastmail.fm to sci.math on Tue Feb 17 11:38:48 2026
    From Newsgroup: sci.math

    Hi,

    Geoffrey E. Hinton, the Nobel Prize winner
    for AI. He was already beating the drums
    for ReLU in 2010:

    HRectified Linear Units Improve Restricted Boltzmann Machines
    Geoffrey E. Hinton & Vinod Nair - 2010 https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf

    Because ANNs (Artificial Neural Networks) were originally
    designed with other functions, e.g. with Logistic function:

    An artificial neuron is a mathematical function conceived
    as a model of a biological neuron in a neural network. https://en.wikipedia.org/wiki/Artificial_neuron

    If you populate additive factor graphs with log P,
    you basically get multiplicative factor graphs.
    So an ANN can express belief networks, right?

    Bye

    P.S.: What is all the hype about Causal AI, and
    the Ladder of Causation |a la Judea Pearl?

    Causal AI rCo the next gen AI
    Prof. Sotirios A. Tsaftaris - 2025
    https://www.youtube.com/watch?v=IelslFzdsYw
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.math on Thu Feb 19 20:02:59 2026
    From Newsgroup: sci.math

    Hi,

    Looks like the Causal AI viewpoint on neural
    networks makes people comfortable. Some
    work from Pittsburgh:

    Causal Representation Learning and Generative AI
    by Dr Kun Zhang #CausalNeSyAI
    https://www.youtube.com/watch?v=UwcitNXphog

    Kathleen M. Carley: Understanding Influence
    A Network Science + AI Approach rCo IC2S2 2025 Keynote https://www.youtube.com/watch?v=iHjFXtIeXnw

    So Gen Alpha will grow up with Causal AI,
    while Gen Z just botched Scryer Prolog, its
    definitively a dead horse now?

    Bye

    Mild Shock schrieb:
    Hi,

    Geoffrey E. Hinton, the Nobel Prize winner
    for AI. He was already beating the drums
    for ReLU in 2010:

    HRectified Linear Units Improve Restricted Boltzmann Machines
    Geoffrey E. Hinton & Vinod Nair - 2010 https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf

    Because ANNs (Artificial Neural Networks) were originally
    designed with other functions, e.g. with Logistic function:

    An artificial neuron is a mathematical function conceived
    as a model of a biological neuron in a neural network. https://en.wikipedia.org/wiki/Artificial_neuron

    If you populate additive factor graphs with log P,
    you basically get multiplicative factor graphs.
    So an ANN can express belief networks, right?

    Bye

    P.S.: What is all the hype about Causal AI, and
    the Ladder of Causation |a la Judea Pearl?

    Causal AI rCo the next gen AI
    Prof. Sotirios A. Tsaftaris - 2025 https://www.youtube.com/watch?v=IelslFzdsYw

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.math on Fri Feb 27 10:33:26 2026
    From Newsgroup: sci.math

    Hi,

    Large Reasoning Models (LRM) seem to move from
    Foundation of Mathematics (FOM) to Theoretical
    Computer Science (TCS). FOM typically gives

    you "white science" mathematics, with sets and
    infinity, if you are lucky a little recursion
    theory. Fun fact TCS is even more "white".

    Interesting paper in as far:

    Lean Meets Theoretical Computer Science:
    Scalable Synthesis of Theorem Proving Challenges
    in Formal-Informal Pairs
    Terry Jingchen Zhang et. al. - 2025
    https://arxiv.org/abs/2508.15878v1

    One swallow does not make a summer?

    But its probably a necessary step. The above
    paper using Busy Beaver and Interger Constraints
    as examples. What logical frameworks do even

    apply, is it enough to have a "total function"
    theory layer, or does TCS need more. TCS can
    be heavy on all sort of discrete and

    non-discrete mathematics.

    Bye

    Mild Shock schrieb:
    Hi,

    Geoffrey E. Hinton, the Nobel Prize winner
    for AI. He was already beating the drums
    for ReLU in 2010:

    HRectified Linear Units Improve Restricted Boltzmann Machines
    Geoffrey E. Hinton & Vinod Nair - 2010 https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf

    Because ANNs (Artificial Neural Networks) were originally
    designed with other functions, e.g. with Logistic function:

    An artificial neuron is a mathematical function conceived
    as a model of a biological neuron in a neural network. https://en.wikipedia.org/wiki/Artificial_neuron

    If you populate additive factor graphs with log P,
    you basically get multiplicative factor graphs.
    So an ANN can express belief networks, right?

    Bye

    P.S.: What is all the hype about Causal AI, and
    the Ladder of Causation |a la Judea Pearl?

    Causal AI rCo the next gen AI
    Prof. Sotirios A. Tsaftaris - 2025 https://www.youtube.com/watch?v=IelslFzdsYw

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to sci.math on Fri Feb 27 09:55:21 2026
    From Newsgroup: sci.math

    On 02/27/2026 01:33 AM, Mild Shock wrote:
    Hi,

    Large Reasoning Models (LRM) seem to move from
    Foundation of Mathematics (FOM) to Theoretical
    Computer Science (TCS). FOM typically gives

    you "white science" mathematics, with sets and
    infinity, if you are lucky a little recursion
    theory. Fun fact TCS is even more "white".

    Interesting paper in as far:

    Lean Meets Theoretical Computer Science:
    Scalable Synthesis of Theorem Proving Challenges
    in Formal-Informal Pairs
    Terry Jingchen Zhang et. al. - 2025
    https://arxiv.org/abs/2508.15878v1

    One swallow does not make a summer?

    But its probably a necessary step. The above
    paper using Busy Beaver and Interger Constraints
    as examples. What logical frameworks do even

    apply, is it enough to have a "total function"
    theory layer, or does TCS need more. TCS can
    be heavy on all sort of discrete and

    non-discrete mathematics.

    Bye

    Mild Shock schrieb:
    Hi,

    Geoffrey E. Hinton, the Nobel Prize winner
    for AI. He was already beating the drums
    for ReLU in 2010:

    HRectified Linear Units Improve Restricted Boltzmann Machines
    Geoffrey E. Hinton & Vinod Nair - 2010 https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf

    Because ANNs (Artificial Neural Networks) were originally
    designed with other functions, e.g. with Logistic function:

    An artificial neuron is a mathematical function conceived
    as a model of a biological neuron in a neural network. https://en.wikipedia.org/wiki/Artificial_neuron

    If you populate additive factor graphs with log P,
    you basically get multiplicative factor graphs.
    So an ANN can express belief networks, right?

    Bye

    P.S.: What is all the hype about Causal AI, and
    the Ladder of Causation |a la Judea Pearl?

    Causal AI rCo the next gen AI
    Prof. Sotirios A. Tsaftaris - 2025 https://www.youtube.com/watch?v=IelslFzdsYw


    "Old-wrapped-as-new" combined with "reinventing-the-wrong".

    It's either equi-interpretable with the rest of mathematics
    like model theory is equi-interpretable with proof theory,
    or it's not.

    Why don't you go "auto" yourself into perfect agreement,
    pretty yourself up in the mirror. That Microsoft's "Lean"
    has a weasel-word in its language, about which its presence
    can be used to derive ready contradictions.


    I don't get what you mean by "white". You mean like
    the "ivory tower"? See I thought "set theory" was
    kind of inclusive ethnically by bringing in "Aleph"
    numbers, while though Greek, Latin, and Hebrew are
    all kind of Mediterranean, that generally being
    considered "olive". Of course the usual digits
    are Arabic (or rather, Indic). wa shi chur li u
    and so on.

    Otherwise I think what you got there is
    "jealousy lime".


    How about "non-standard" mathematics, then.


    Anyways "descriptive set theory" after "set theory:
    a theory of one relation" is pretty usual.

    That said there are other theories of one relation,
    like class theory or ordering theory about class/set
    distinction and cardinal/ordinal distinction and
    part/particle distinction and so on, and the primary
    objects in geometry and number theory aren't exactly
    each other without infinite completions and super-classical
    reasoning and careful book-keeping of relations of rates
    of resources in mathematics, then about arithmetic and
    algebra and analysis their calculi, given function theory
    and topology then an operator calculus about that.

    So, a "heno-theory" rather results then for a "mono-heno-theory".



    It's sort of like the story of Differential Geometry, "DG".
    "Hi DG, what's happenin'." "Well you know classical continuous
    functions, with no infinite asymptotes." "Sure, why not."
    "Well we've found a restriction of those to where there's also
    no zero tangents and thusly derived a bunch of results that
    basically we stole from Gregory but mostly Coates yet among
    others that it results a bunch of closed forms of analytical
    expressions in terms of these functions about immersions and
    submersions in differential geometry." "Wow, fantastic, DG,
    that's a usual set of results in those DG-functions." "Oh,
    no, we restricted the comprehension that it's a restriction
    of comprehension, and claim those are classical continuous
    functions now about that thusly that's the definition of
    function." "Well, DG, I guess all you have to do is re-write
    all the curriculum in the world and blot out the rest and
    claim yourself king."

    What results isn't quite all "mathematics", ...,
    yet both your monitors as one screen agree.


    How about "Mizar".


    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to sci.math on Fri Feb 27 10:57:49 2026
    From Newsgroup: sci.math

    On 02/27/2026 09:55 AM, Ross Finlayson wrote:
    On 02/27/2026 01:33 AM, Mild Shock wrote:
    Hi,

    Large Reasoning Models (LRM) seem to move from
    Foundation of Mathematics (FOM) to Theoretical
    Computer Science (TCS). FOM typically gives

    you "white science" mathematics, with sets and
    infinity, if you are lucky a little recursion
    theory. Fun fact TCS is even more "white".

    Interesting paper in as far:

    Lean Meets Theoretical Computer Science:
    Scalable Synthesis of Theorem Proving Challenges
    in Formal-Informal Pairs
    Terry Jingchen Zhang et. al. - 2025
    https://arxiv.org/abs/2508.15878v1

    One swallow does not make a summer?

    But its probably a necessary step. The above
    paper using Busy Beaver and Interger Constraints
    as examples. What logical frameworks do even

    apply, is it enough to have a "total function"
    theory layer, or does TCS need more. TCS can
    be heavy on all sort of discrete and

    non-discrete mathematics.

    Bye

    Mild Shock schrieb:
    Hi,

    Geoffrey E. Hinton, the Nobel Prize winner
    for AI. He was already beating the drums
    for ReLU in 2010:

    HRectified Linear Units Improve Restricted Boltzmann Machines
    Geoffrey E. Hinton & Vinod Nair - 2010
    https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf

    Because ANNs (Artificial Neural Networks) were originally
    designed with other functions, e.g. with Logistic function:

    An artificial neuron is a mathematical function conceived
    as a model of a biological neuron in a neural network.
    https://en.wikipedia.org/wiki/Artificial_neuron

    If you populate additive factor graphs with log P,
    you basically get multiplicative factor graphs.
    So an ANN can express belief networks, right?

    Bye

    P.S.: What is all the hype about Causal AI, and
    the Ladder of Causation |a la Judea Pearl?

    Causal AI rCo the next gen AI
    Prof. Sotirios A. Tsaftaris - 2025
    https://www.youtube.com/watch?v=IelslFzdsYw


    "Old-wrapped-as-new" combined with "reinventing-the-wrong".

    It's either equi-interpretable with the rest of mathematics
    like model theory is equi-interpretable with proof theory,
    or it's not.

    Why don't you go "auto" yourself into perfect agreement,
    pretty yourself up in the mirror. That Microsoft's "Lean"
    has a weasel-word in its language, about which its presence
    can be used to derive ready contradictions.


    I don't get what you mean by "white". You mean like
    the "ivory tower"? See I thought "set theory" was
    kind of inclusive ethnically by bringing in "Aleph"
    numbers, while though Greek, Latin, and Hebrew are
    all kind of Mediterranean, that generally being
    considered "olive". Of course the usual digits
    are Arabic (or rather, Indic). wa shi chur li u
    and so on.

    Otherwise I think what you got there is
    "jealousy lime".


    How about "non-standard" mathematics, then.


    Anyways "descriptive set theory" after "set theory:
    a theory of one relation" is pretty usual.

    That said there are other theories of one relation,
    like class theory or ordering theory about class/set
    distinction and cardinal/ordinal distinction and
    part/particle distinction and so on, and the primary
    objects in geometry and number theory aren't exactly
    each other without infinite completions and super-classical
    reasoning and careful book-keeping of relations of rates
    of resources in mathematics, then about arithmetic and
    algebra and analysis their calculi, given function theory
    and topology then an operator calculus about that.

    So, a "heno-theory" rather results then for a "mono-heno-theory".



    It's sort of like the story of Differential Geometry, "DG".
    "Hi DG, what's happenin'." "Well you know classical continuous
    functions, with no infinite asymptotes." "Sure, why not."
    "Well we've found a restriction of those to where there's also
    no zero tangents and thusly derived a bunch of results that
    basically we stole from Gregory but mostly Coates yet among
    others that it results a bunch of closed forms of analytical
    expressions in terms of these functions about immersions and
    submersions in differential geometry." "Wow, fantastic, DG,
    that's a usual set of results in those DG-functions." "Oh,
    no, we restricted the comprehension that it's a restriction
    of comprehension, and claim those are classical continuous
    functions now about that thusly that's the definition of
    function." "Well, DG, I guess all you have to do is re-write
    all the curriculum in the world and blot out the rest and
    claim yourself king."

    What results isn't quite all "mathematics", ...,
    yet both your monitors as one screen agree.


    How about "Mizar".



    It's like when Modi visited the Knesset the other day
    and reeled off a bunch of diplomatic platitudes then
    mentioned "The Gazans have been mistreated terribly
    and that's wrong." Then the news spin was a bunch of
    Fudd Bibi with "I have friends who eat curry."
    FB-functions.

    Not to just start making jokes about Jews since it's a
    usual thing that only people of a given ethnicity can
    perpetuate ethnic stereotypes without being in worse bad taste,
    and the only way I could theoretically be a Jew is
    matrilineally, but if you make a silhouette of Argentina's
    Milei, the president of Hungary, Fudd Bibi, and the
    Penguin from Batman, it's sort of one outline and quacks.

    I suppose though that's jokes about Russo-Hungarians
    and Batman villains. Or fat jokes, ....

    It's like, "what's the difference between
    the IDF and the RSF?" "Oh, about 350 miles."



    (I'm more of a Czech Pole, about 1/32. I asked Grandma
    one time what we were and she said "Germans", and "Methodists".
    I don't recall any names that end with vowels or begin
    with 'O' except "Peplau", and "Thebo". I don't have
    anything with "skulls" on them yet lots with "eagles".)

    "Boy, Superman, you sure are an asshole when you're drunk."


    https://en.wikipedia.org/wiki/Gracie_family


    December 10 is the anniversary of the United Nations
    Universal Declaration of Human Rights, which was a
    body formed to prevent the recurrence of atrocities.
    Richard McKeon was one of its authors, though he's
    better known for a great account of "Aristotle".


    These times, ....

    It's like how big a greaseball is Bubba,
    I don't know, but I don't want him feeling my pain.
    "I'm not saying she's a gold-digger, ...".


    It's like Trump and Putin were visiting and driving
    around in a limousine and getting high and get to
    boasting. Trump says "my guards will do anything I say"
    and Putin says "my guards will do anything I say." So
    Trump calls over one of his guards and says "Jump off
    this bridge." The guard says "No can do Mr. President,
    I have a wife and three kids at home. If I died there
    would be no-one to take care of them." Trump says
    "You're fired and black-balled and if you ever breathe
    a word of this I'll sue you for a jazillion dollars."
    Putin calls over one his guards and says "Jump off
    this bridge." The glum guard says "Da, Mr. President."
    Trump starts fuming and spitting and says "Why would
    you do that?" And the glum guard says, "Well, you see,
    Orange Fat Blowhard, I too have a wife and three kids at home.
    If I don't die there will be no-one to take care of me."


    It's like when Trump was on a bribe junket plane crossing
    the ocean and the engines go out and it's going to crash.
    Say there are no parachutes or one of his goblins took them
    all. So Trump gets to waling and starts praying "If anybody's
    listening please save us." All of a sudden he hears a guy
    gently clearing his throat and across the aisle is sitting
    a man of some apparent wealth and taste. Trumps cries
    "It's a miracle, it must be an angel, save us!". And the guy
    says "Actually I'm Satan." Then Trump says "Let's make a deal,
    I hear you do deals, I'll do anything." Then Satan shakes his
    head and pulls out a sheet of paper with an inky scrawl on
    the dotted line. "Trump, you nincompoop, I'm here to collect."









    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Beauford Barzetti@etua@reraaade.it to sci.physics.relativity,sci.math on Fri Feb 27 21:48:51 2026
    From Newsgroup: sci.math

    Ross Finlayson wrote:

    It's like when Modi visited the Knesset the other day and reeled off a
    bunch of diplomatic platitudes then mentioned "The Gazans have been mistreated terribly and that's wrong." Then the news spin was a bunch of
    Fudd Bibi with "I have friends who eat curry." FB-functions.

    nonsense, the khazar goys are from India, that's where they are coming
    from, first the Great Russia, then the shit europe
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to sci.physics.relativity,sci.math on Fri Feb 27 14:26:42 2026
    From Newsgroup: sci.math

    On 02/27/2026 01:48 PM, Beauford Barzetti wrote:
    Ross Finlayson wrote:

    It's like when Modi visited the Knesset the other day and reeled off a
    bunch of diplomatic platitudes then mentioned "The Gazans have been
    mistreated terribly and that's wrong." Then the news spin was a bunch of
    Fudd Bibi with "I have friends who eat curry." FB-functions.

    nonsense, the khazar goys are from India, that's where they are coming
    from, first the Great Russia, then the shit europe


    There's quite a big difference among Hindus among
    those who swipe the three fingers horizontally, or vertically.

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Marcello =?iso-8859-1?q?Fern=E1ndez?=@rfvea@rlzonful.es to sci.physics.relativity,sci.math on Fri Feb 27 22:56:59 2026
    From Newsgroup: sci.math

    Ross Finlayson wrote:

    On 02/27/2026 01:48 PM, Beauford Barzetti wrote:
    Ross Finlayson wrote:

    It's like when Modi visited the Knesset the other day and reeled off a
    bunch of diplomatic platitudes then mentioned "The Gazans have been
    mistreated terribly and that's wrong." Then the news spin was a bunch
    of Fudd Bibi with "I have friends who eat curry." FB-functions.

    nonsense, the khazar goys are from India, that's where they are coming
    from, first the Great Russia, then the shit europe

    There's quite a big difference among Hindus among those who swipe the
    three fingers horizontally, or vertically.

    hmm, that must be the european harsh colder climate, painting their hair yellow, but they are of Indian originate, also known as indo-european, not european.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to sci.physics.relativity,sci.math on Fri Feb 27 19:59:44 2026
    From Newsgroup: sci.math

    On 02/27/2026 02:56 PM, Marcello Fern|indez wrote:
    Ross Finlayson wrote:

    On 02/27/2026 01:48 PM, Beauford Barzetti wrote:
    Ross Finlayson wrote:

    It's like when Modi visited the Knesset the other day and reeled off a >>>> bunch of diplomatic platitudes then mentioned "The Gazans have been
    mistreated terribly and that's wrong." Then the news spin was a bunch
    of Fudd Bibi with "I have friends who eat curry." FB-functions.

    nonsense, the khazar goys are from India, that's where they are coming
    from, first the Great Russia, then the shit europe

    There's quite a big difference among Hindus among those who swipe the
    three fingers horizontally, or vertically.

    hmm, that must be the european harsh colder climate, painting their hair yellow, but they are of Indian originate, also known as indo-european, not european.


    No, it's actually more "nice" and "mean",
    though I forget which is which and
    don't presume to judge character.


    --- Synchronet 3.21d-Linux NewsLink 1.2
  • From Levy Stamatas@tl@ayetl.gr to sci.physics.relativity,sci.math on Sat Feb 28 12:49:11 2026
    From Newsgroup: sci.math

    Ross Finlayson wrote:

    hmm, that must be the european harsh colder climate, painting their
    hair yellow, but they are of Indian originate, also known as
    indo-european, not european.


    No, it's actually more "nice" and "mean",
    though I forget which is which and don't presume to judge character.

    what the hell is talking about this wanker, never heard of indo-european.
    The khazar goys wannabe jews are blacks from india invading europe,
    shitting without toilet paper as we speak. That's a written old tradition.
    --- Synchronet 3.21d-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to sci.physics.relativity,sci.math on Sat Feb 28 06:47:39 2026
    From Newsgroup: sci.math

    On 02/28/2026 04:49 AM, Levy Stamatas wrote:
    Ross Finlayson wrote:

    hmm, that must be the european harsh colder climate, painting their
    hair yellow, but they are of Indian originate, also known as
    indo-european, not european.


    No, it's actually more "nice" and "mean",
    though I forget which is which and don't presume to judge character.

    what the hell is talking about this wanker, never heard of indo-european.
    The khazar goys wannabe jews are blacks from india invading europe,
    shitting without toilet paper as we speak. That's a written old tradition.


    Kind of like the difference between Confucians and the Shang.


    Of course I'm aware of significant influences of the Indic
    with regards to the Vedic for example about something like
    P.T. Raju's "The Structural Depth of Indian Thought" and
    after something like Campbell's world histories, the Noachic,
    and Atlantean, and also about a general overview of world events
    circa 200BC - 200AD.

    Then that this includes various sources and pockets of humanity
    is a bit different than usual accounts of inbreeding.


    --- Synchronet 3.21d-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.math on Sat Feb 28 16:33:24 2026
    From Newsgroup: sci.math

    Hi,

    Now "white science" might have more leaning
    towards Causal AI. But sometimes I have the
    feeling Generative AI is the "yellow science"

    now. DeepSeek left handedly inventing the
    notation M(a,b) for matrix representation of
    a dual number a + b e:

    M(a,b) = a I + b E

    And then discussing variants like:

    E = [[ 0 1 ]
    [ 0 0 ]]

    Or this variant like:

    E = [[ 1 -1 ]
    [ 1 -1 ]]

    I am in mild shock! Then I am reading the small print:

    Prof. Cai Wen, Institute of Extenics and Innovative Methods, Guangzhou, https://fs.unm.edu/DualNumbers.pdf

    What is Extenics? Any ideas?

    Have Fun!

    Bye

    Mild Shock schrieb:
    Hi,

    Large Reasoning Models (LRM) seem to move from
    Foundation of Mathematics (FOM) to Theoretical
    Computer Science (TCS). FOM typically gives

    you "white science" mathematics, with sets and
    infinity, if you are lucky a little recursion
    theory. Fun fact TCS is even more "white".

    Interesting paper in as far:

    Lean Meets Theoretical Computer Science:
    Scalable Synthesis of Theorem Proving Challenges
    in Formal-Informal Pairs
    Terry Jingchen Zhang et. al. - 2025
    https://arxiv.org/abs/2508.15878v1

    One swallow does not make a summer?

    But its probably a necessary step. The above
    paper using Busy Beaver and Interger Constraints
    as examples. What logical frameworks do even

    apply, is it enough to have a "total function"
    theory layer, or does TCS need more. TCS can
    be heavy on all sort of discrete and

    non-discrete mathematics.

    Bye

    Mild Shock schrieb:
    Hi,

    Geoffrey E. Hinton, the Nobel Prize winner
    for AI. He was already beating the drums
    for ReLU in 2010:

    HRectified Linear Units Improve Restricted Boltzmann Machines
    Geoffrey E. Hinton & Vinod Nair - 2010 https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf

    Because ANNs (Artificial Neural Networks) were originally
    designed with other functions, e.g. with Logistic function:

    An artificial neuron is a mathematical function conceived
    as a model of a biological neuron in a neural network. https://en.wikipedia.org/wiki/Artificial_neuron

    If you populate additive factor graphs with log P,
    you basically get multiplicative factor graphs.
    So an ANN can express belief networks, right?

    Bye

    P.S.: What is all the hype about Causal AI, and
    the Ladder of Causation |a la Judea Pearl?

    Causal AI rCo the next gen AI
    Prof. Sotirios A. Tsaftaris - 2025 https://www.youtube.com/watch?v=IelslFzdsYw


    --- Synchronet 3.21d-Linux NewsLink 1.2