There are more and more papers of this sort:

Reliable Reasoning Beyond Natural Language
To address this, we propose a neurosymbolic
approach that prompts LLMs to extract and encode
all relevant information from a problem statement as
logical code statements, and then use a logic programming
language (Prolog) to conduct the iterative computations of
explicit deductive reasoning.
[2407.11373] Reliable Reasoning Beyond Natural Language

The future of Prolog is bright?

Mild Shock schrieb:
> I can run the simpler:
>
> x^3 + y^3 + 9 = z^3
>
> In a web browser with Dogelog Player:
>
> Example 71: Diophantine Modular
> X = 216, Y = 52, Z = 217;
> X = 52, Y = 216, Z = 217;
> fail.
> % Zeit 3574 ms, GC 7 ms, Lips 1696084, Uhr 01.09.2023 20:56 true.
> https://jsfiddle.net/Jean_Luc_Picard_2021/d2njehtp
>
> Since it is a different algorithm than the usual Prolog
> CLP(FD), its faster than SWI-Prolog and Scryer Prolog,
> even faster than desktop.
>
> Have Fun!
>
> P.S.: Unfortunately I couldn't cross yet the 42 barrier.
>
> Mostowski Collapse schrieb am Dienstag, 20. September 2022 um 09:37:53 UTC+2:
>> Well I mean:
>>
>> x^3 + y^3 + 42 = z^3
>>
>> You only need to consider positive x,y,z then.
>> Assume there is some ingenious approach to
>> speed it up, maybe could then also solve:
>>
>> 114, 390, 627, 633, 732, 921, 975.
>> Nasser M. Abbasi schrieb am Dienstag, 20. September 2022 um 04:24:19 UTC+2:
>>> On 9/19/2022 4:51 PM, Mostowski Collapse wrote:
>>>> Rumors are, that finding an integer solution
>>>> took a lot of computing resources:
>>>>
>>>> x^3 + y^3 + 42 = z^3
>>>>
>>>> Can we nevertheless solve this on a PC in
>>>> less than 5 minutes?
>>> Did you mean
>>>
>>> x^3 + y^3 + z^3 = 42?
>>>
>>> Given that it says at
>>> <https://interestingengineering.com/science/the-sum-of-three-cubes-problem-for-42-has-just-been-solved>
>>>
>>> "Charity Engine uses a computer's idle processing power, and it took 1 million
>>> hours of processing time to solve the Diophantine Equation where
>>> k is equal to 42. The answer is:
>>> (-80538738812075974)^3 + (80435758145817515)^3 + (12602123297335631)^3 = 42."
>>>
>>> I do not see how this can be found in 5 minutes on a PC? No matter
>>> how fast the PC is. They seem to have done brute force search to find this.
>>> But may be they used improved search method as given in
>>>
>>> <https://news.mit.edu/2021/solution-3-sum-cubes-puzzle-0311>
>>> under "A solution’s twist" section. But even that will stil require
>>> very large brute search method.
>>>
>>> But I tried
>>>
>>> eq=x^3+y^3+42==z^3
>>> FindInstance[eq&&Element[{x,y,z},Integers],{x,y,z},1]
>>>
>>> And it said
>>>
>>> FindInstance::nsmet: The methods available to FindInstance are insufficient to
>>> find the requested instances or prove they do not exist.
>>>
>>> in 0.001 seconds. I guess Mathematica still does not know about
>>> the method used to solve for 42.
>>>
>>> --Nasser

Mild Shock schrieb:
>
> There are more and more papers of this sort:
>
> Reliable Reasoning Beyond Natural Language
> To address this, we propose a neurosymbolic
> approach that prompts LLMs to extract and encode
> all relevant information from a problem statement as
> logical code statements, and then use a logic programming
> language (Prolog) to conduct the iterative computations of
> explicit deductive reasoning.
> [2407.11373] Reliable Reasoning Beyond Natural Language
>
> The future of Prolog is bright?
>
> [...]

Thanks for the message. Unfortunately, <sci.math.symbolic> is no longer
archived at <narkive.com>, the last post shown there being 4 months old
already (posts were anyway shown only if they received at least one
answer). Newsgroups not archived and not crawled by search engines are
doomed, I think.

Martin.

What news provider are you using.
I am using solani.org. I see all

messages they are still there:

- 04.08.2024, 22:12 Mild Shock
- 11.08.2024, 23:10 clicliclic@freenet.de

Also I don't mind if they are not
indefinitely archived, LLM is

anyway a moving target. So who cares?
Just use USENET for news and

discussion as in the old times before
the ggexit (google groups exit).

LoL

clicliclic@freenet.de schrieb:
>
> Mild Shock schrieb:
>>
>> There are more and more papers of this sort:
>>
>> Reliable Reasoning Beyond Natural Language
>> To address this, we propose a neurosymbolic
>> approach that prompts LLMs to extract and encode
>> all relevant information from a problem statement as
>> logical code statements, and then use a logic programming
>> language (Prolog) to conduct the iterative computations of
>> explicit deductive reasoning.
>> [2407.11373] Reliable Reasoning Beyond Natural Language
>>
>> The future of Prolog is bright?
>>
>> [...]
>
> Thanks for the message. Unfortunately, <sci.math.symbolic> is no longer
> archived at <narkive.com>, the last post shown there being 4 months old
> already (posts were anyway shown only if they received at least one
> answer). Newsgroups not archived and not crawled by search engines are
> doomed, I think.
>
> Martin.
>

Mild Shock schrieb:
>
> clicliclic@freenet.de schrieb:
> >
> > Mild Shock schrieb:
> >>
> >> There are more and more papers of this sort:
> >>
> >> Reliable Reasoning Beyond Natural Language
> >> To address this, we propose a neurosymbolic
> >> approach that prompts LLMs to extract and encode
> >> all relevant information from a problem statement as
> >> logical code statements, and then use a logic programming
> >> language (Prolog) to conduct the iterative computations of
> >> explicit deductive reasoning.
> >> [2407.11373] Reliable Reasoning Beyond Natural Language
> >>
> >> The future of Prolog is bright?
> >>
> >> [...]
> >
> > Thanks for the message. Unfortunately, <sci.math.symbolic> is no
> > longer archived at <narkive.com>, the last post shown there being 4
> > months old already (posts were anyway shown only if they received
> > at least one answer). Newsgroups not archived and not crawled by
> > search engines are doomed, I think.
> >
>
> What news provider are you using.
> I am using solani.org. I see all
>
> messages they are still there:
>
> - 04.08.2024, 22:12 Mild Shock
> - 11.08.2024, 23:10 clicliclic@freenet.de
>
> Also I don't mind if they are not
> indefinitely archived, LLM is
>
> anyway a moving target. So who cares?
> Just use USENET for news and
>
> discussion as in the old times before
> the ggexit (google groups exit).
>

Yes, but ... there were multiple Usenet archives before Google Groups
started, some reaching back to the beginnings of Usenet; Google bought
one of them and the competitors subsequently decided to shut down. Only
<narkive.com> seems to be left now, but doesn't seem to function
reliably. Somebody should alert them, perhaps.

Message retention at the Usenet providers themselves ranges from a few
months to many years. In the past, I found Eternal September to have a
good retention. Apply at:

<https://www.eternal-september.org/>

and point your newreader to:

<news://news.eternal-september.org>

when you got a password.

Martin.

Maybe you don't undertsand USENET.
But I don't need to register with
eternal-september , I already

registered with solana. You can
check yourself the posts I did
via solana are also visible

in eternal-september . Just go
check this fact yourself please.

clicliclic@freenet.de schrieb:
> Message retention at the Usenet providers themselves ranges from a few
> months to many years. In the past, I found Eternal September to have a
> good retention. Apply at:
>
> <https://www.eternal-september.org/>
>
> and point your newreader to:
>
> <news://news.eternal-september.org>
>
> when you got a password.
>
> Martin.
>

Corr: solana --> solani

Do you think messages posted via
eternal-september have a longer retention,
than messages posted via solana, that

also arrive in eternal-september ?
I don't know, and I still opininate
that its not relevant, since for

example LLM is a quickly moving
technology. But lets say somebody
is really interested in retention time,

what do we know about it?

- solani has a few groups which are held
indefinitely, and other groups are held
only 4 months.

- eternal-september has a few groups which
are held 2 years, but then it can go
down to 3 months.

Mild Shock schrieb:
>
> Maybe you don't undertsand USENET.
> But I don't need to register with
> eternal-september , I already
>
> registered with solana. You can
> check yourself the posts I did
> via solana are also visible
>
> in eternal-september . Just go
> check this fact yourself please.
>
> clicliclic@freenet.de schrieb:
>> Message retention at the Usenet providers themselves ranges from a few
>> months to many years. In the past, I found Eternal September to have a
>> good retention. Apply at:
>>
>> <https://www.eternal-september.org/>
>>
>> and point your newreader to:
>>
>> <news://news.eternal-september.org>
>>
>> when you got a password.
>>
>> Martin.
>>
>

Ha Ha, I did the below post
with eternal-september, and now
I am using solani again.

But thx, I now tree news server
registered, you can view them all
at once for example with Seamonkey,

news.solani.org
news.gmane.io
news.eternal-september.org

Its always convenient to have multiple
news server at hand, because not
all newsgroups are available in all

news servers.

j4nbur53 schrieb:
> Corr: solana --> solani
>
> Do you think messages posted via
> eternal-september have a longer retention,
> than messages posted via solana, that
>
> also arrive in eternal-september ?
> I don't know, and I still opininate
> that its not relevant, since for
>
> example LLM is a quickly moving
> technology. But lets say somebody
> is really interested in retention time,
>
> what do we know about it?
>
> -  solani has a few groups which are held
> indefinitely, and other groups are held
> only 4 months.
>
> - eternal-september has a few groups which
> are held 2 years, but then it can go
> down to 3 months.
>
> Mild Shock schrieb:
>>
>> Maybe you don't undertsand USENET.
>> But I don't need to register with
>> eternal-september , I already
>>
>> registered with solana. You can
>> check yourself the posts I did
>> via solana are also visible
>>
>> in eternal-september . Just go
>> check this fact yourself please.
>>
>> clicliclic@freenet.de schrieb:
>>> Message retention at the Usenet providers themselves ranges from a few
>>> months to many years. In the past, I found Eternal September to have a
>>> good retention. Apply at:
>>>
>>> <https://www.eternal-september.org/>
>>>
>>> and point your newreader to:
>>>
>>> <news://news.eternal-september.org>
>>>
>>> when you got a password.
>>>
>>> Martin.
>>>
>>
>

An example group in eternal-september,
but not in solani and not in gmane

is for example: alt.drugs.usenet

But it doesn't have any content, :-(
So maybe eternal-september is

less moderated? I don't know.

Mild Shock schrieb:
> Ha Ha, I did the below post
> with eternal-september, and now
> I am using solani again.
>
> But thx, I now tree news server
> registered, you can view them all
> at once for example with Seamonkey,
>
> news.solani.org
> news.gmane.io
> news.eternal-september.org
>
> Its always convenient to have multiple
> news server at hand, because not
> all newsgroups are available in all
>
> news servers.
>
> j4nbur53 schrieb:
>> Corr: solana --> solani
>>
>> Do you think messages posted via
>> eternal-september have a longer retention,
>> than messages posted via solana, that
>>
>> also arrive in eternal-september ?
>> I don't know, and I still opininate
>> that its not relevant, since for
>>
>> example LLM is a quickly moving
>> technology. But lets say somebody
>> is really interested in retention time,
>>
>> what do we know about it?
>>
>> -  solani has a few groups which are held
>> indefinitely, and other groups are held
>> only 4 months.
>>
>> - eternal-september has a few groups which
>> are held 2 years, but then it can go
>> down to 3 months.
>>
>> Mild Shock schrieb:
>>>
>>> Maybe you don't undertsand USENET.
>>> But I don't need to register with
>>> eternal-september , I already
>>>
>>> registered with solana. You can
>>> check yourself the posts I did
>>> via solana are also visible
>>>
>>> in eternal-september . Just go
>>> check this fact yourself please.
>>>
>>> clicliclic@freenet.de schrieb:
>>>> Message retention at the Usenet providers themselves ranges from a few
>>>> months to many years. In the past, I found Eternal September to have a
>>>> good retention. Apply at:
>>>>
>>>> <https://www.eternal-september.org/>
>>>>
>>>> and point your newreader to:
>>>>
>>>> <news://news.eternal-september.org>
>>>>
>>>> when you got a password.
>>>>
>>>> Martin.
>>>>
>>>
>>
>

Now I wonder whether LLMs should be an
inch more informed by results from Neuro-
endocrinology research. I remember Marvin
Minsky publishing his ‘The Society of Mind’:

Introduction to ‘The Society of Mind’
https://www.youtube.com/watch?v=-pb3z2w9gDg

But this made me think about a multi agent
systems. Now with LLMs what about a new
connectionist and deep learning approach.
Plus Prolog for the pre frontal cortex (PFC).

But who can write a blue print? Now there
is this amazing guy called Robert M. Sapolsky
who recently published Determined: A Science
of Life without Free Will, who

calls consciousness just a hicup. His turtles
all the way down model is a tour de force
through an unsettling conclusion: We may not
grasp the precise marriage of nature and nurture

that creates the physics and chemistry at the
base of human behavior, but that doesn’t mean it
doesn’t exist. But the pre frontal cortex (PFC)
seems to be still quite brittle and not extremly

performant and quite energy hungry.
So Prolog might excell?

Determined: A Science of Life Without Free Will
https://www.amazon.de/dp/0525560998

Mild Shock schrieb:
>
> There are more and more papers of this sort:
>
> Reliable Reasoning Beyond Natural Language
> To address this, we propose a neurosymbolic
> approach that prompts LLMs to extract and encode
> all relevant information from a problem statement as
> logical code statements, and then use a logic programming
> language (Prolog) to conduct the iterative computations of
> explicit deductive reasoning.
> [2407.11373] Reliable Reasoning Beyond Natural Language
>
> The future of Prolog is bright?
>
> Mild Shock schrieb:
>> I can run the simpler:
>>
>> x^3 + y^3 + 9 = z^3
>>
>> In a web browser with Dogelog Player:
>>
>> Example 71: Diophantine Modular
>> X = 216, Y = 52, Z = 217;
>> X = 52, Y = 216, Z = 217;
>> fail.
>> % Zeit 3574 ms, GC 7 ms, Lips 1696084, Uhr 01.09.2023 20:56 true.
>> https://jsfiddle.net/Jean_Luc_Picard_2021/d2njehtp
>>
>> Since it is a different algorithm than the usual Prolog
>> CLP(FD), its faster than SWI-Prolog and Scryer Prolog,
>> even faster than desktop.
>>
>> Have Fun!
>>
>> P.S.: Unfortunately I couldn't cross yet the 42 barrier.
>>
>> Mostowski Collapse schrieb am Dienstag, 20. September 2022 um 09:37:53
>> UTC+2:
>>> Well I mean:
>>>
>>> x^3 + y^3 + 42 = z^3
>>>
>>> You only need to consider positive x,y,z then.
>>> Assume there is some ingenious approach to
>>> speed it up, maybe could then also solve:
>>>
>>> 114, 390, 627, 633, 732, 921, 975.
>>> Nasser M. Abbasi schrieb am Dienstag, 20. September 2022 um 04:24:19
>>> UTC+2:
>>>> On 9/19/2022 4:51 PM, Mostowski Collapse wrote:
>>>>> Rumors are, that finding an integer solution
>>>>> took a lot of computing resources:
>>>>>
>>>>> x^3 + y^3 + 42 = z^3
>>>>>
>>>>> Can we nevertheless solve this on a PC in
>>>>> less than 5 minutes?
>>>> Did you mean
>>>>
>>>> x^3 + y^3 + z^3 = 42?
>>>>
>>>> Given that it says at
>>>> <https://interestingengineering.com/science/the-sum-of-three-cubes-problem-for-42-has-just-been-solved>
>>>>
>>>>
>>>> "Charity Engine uses a computer's idle processing power, and it took
>>>> 1 million
>>>> hours of processing time to solve the Diophantine Equation where
>>>> k is equal to 42. The answer is:
>>>> (-80538738812075974)^3 + (80435758145817515)^3 +
>>>> (12602123297335631)^3 = 42."
>>>>
>>>> I do not see how this can be found in 5 minutes on a PC? No matter
>>>> how fast the PC is. They seem to have done brute force search to
>>>> find this.
>>>> But may be they used improved search method as given in
>>>>
>>>> <https://news.mit.edu/2021/solution-3-sum-cubes-puzzle-0311>
>>>> under "A solution’s twist" section. But even that will stil require
>>>> very large brute search method.
>>>>
>>>> But I tried
>>>>
>>>> eq=x^3+y^3+42==z^3
>>>> FindInstance[eq&&Element[{x,y,z},Integers],{x,y,z},1]
>>>>
>>>> And it said
>>>>
>>>> FindInstance::nsmet: The methods available to FindInstance are
>>>> insufficient to
>>>> find the requested instances or prove they do not exist.
>>>>
>>>> in 0.001 seconds. I guess Mathematica still does not know about
>>>> the method used to solve for 42.
>>>>
>>>> --Nasser
>