Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.theory > #35914 > unrolled thread
| Started by | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| First post | 2021-07-07 18:19 -0700 |
| Last post | 2021-07-30 15:22 +0100 |
| Articles | 20 on this page of 51 — 5 participants |
Back to article view | Back to comp.theory
revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-07 18:19 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-08 02:44 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N DV <xlt.pjw@gmail.com> - 2021-07-08 05:37 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Jeff Barnett <jbb@notatt.com> - 2021-07-08 14:21 -0600
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-14 07:28 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N DV <xlt.pjw@gmail.com> - 2021-07-14 14:28 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-15 12:51 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-18 03:45 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N DV <xlt.pjw@gmail.com> - 2021-07-18 06:48 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-18 14:28 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-26 04:45 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-26 14:35 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 00:35 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 02:40 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-28 17:27 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 11:50 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-28 20:52 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 15:53 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 16:58 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 17:41 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-29 01:45 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 20:07 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-28 20:27 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-29 14:59 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Malcolm McLean <malcolm.arthur.mclean@gmail.com> - 2021-07-29 02:27 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 06:24 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Malcolm McLean <malcolm.arthur.mclean@gmail.com> - 2021-07-29 07:50 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Jeff Barnett <jbb@notatt.com> - 2021-07-29 11:44 -0600
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 11:59 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Jeff Barnett <jbb@notatt.com> - 2021-07-29 13:23 -0600
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-29 20:04 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-29 15:56 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 11:53 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-29 20:12 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 13:12 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-29 23:18 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 15:36 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 16:38 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-30 00:59 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 17:05 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-30 01:19 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 17:27 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-30 02:22 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-29 19:16 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Jeff Barnett <jbb@notatt.com> - 2021-07-29 22:25 -0600
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-30 00:35 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-30 10:53 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-30 05:19 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-30 14:22 +0100
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Daniel Pehoushek <pehoushek1@gmail.com> - 2021-07-30 07:16 -0700
Re: revised relativization theorem: P=NP for small N, P!=NP for large N Ben Bacarisse <ben.usenet@bsb.me.uk> - 2021-07-30 15:22 +0100
Page 1 of 3 [1] 2 3 Next page →
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-07 18:19 -0700 |
| Subject | revised relativization theorem: P=NP for small N, P!=NP for large N |
| Message-ID | <1b5250be-f4d3-4841-aadb-37a818975355n@googlegroups.com> |
simple is good lifting up symbols halfway above the line is bad daniel (little d)
[toc] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-08 02:44 -0700 |
| Message-ID | <97b45bc9-7b03-40cc-a785-06792d6a7ba8n@googlegroups.com> |
| In reply to | #35914 |
On Wednesday, July 7, 2021 at 9:19:10 PM UTC-4, Daniel Pehoushek wrote: > simple is good > lifting up symbols halfway above the line is bad > > daniel (little d) monotone reason is linear. abortion is bad. weapons are bad.
[toc] | [prev] | [next] | [standalone]
| From | DV <xlt.pjw@gmail.com> |
|---|---|
| Date | 2021-07-08 05:37 -0700 |
| Message-ID | <1afd7434-6b14-4ced-82c3-e500742df824n@googlegroups.com> |
| In reply to | #35931 |
On Thursday, July 8, 2021 at 5:44:07 AM UTC-4, pehou...@gmail.com wrote: > On Wednesday, July 7, 2021 at 9:19:10 PM UTC-4, Daniel Pehoushek wrote: > > simple is good > > lifting up symbols halfway above the line is bad > > > > daniel (little d) > monotone reason is linear. > abortion is bad. > weapons are bad. Do you mind if I ask what monotone reason is?
[toc] | [prev] | [next] | [standalone]
| From | Jeff Barnett <jbb@notatt.com> |
|---|---|
| Date | 2021-07-08 14:21 -0600 |
| Message-ID | <sc7mnp$km1$1@dont-email.me> |
| In reply to | #35937 |
On 7/8/2021 6:37 AM, DV wrote: > On Thursday, July 8, 2021 at 5:44:07 AM UTC-4, pehou...@gmail.com wrote: >> On Wednesday, July 7, 2021 at 9:19:10 PM UTC-4, Daniel Pehoushek wrote: >>> simple is good >>> lifting up symbols halfway above the line is bad >>> >>> daniel (little d) >> monotone reason is linear. >> abortion is bad. >> weapons are bad. > > Do you mind if I ask what monotone reason is? A "monotone reasoner" is a system/agent that can learn more but cannot learn different. Example: It is known that "Bob has a car." We can learn that "Bob's car is red." but wee cannot deal with "Bob actually has a skate board, not a car". The issue with non-monotonic reasoning is that inferences made in the past can be invalidated by new knowledge and it's computationally expensive and god-awful hard to back out those now-wrong results. Learning and proving mathematics is in principal monotonic. You can always rely on old theorems and conclusions as long as you are not taught something that is inconsistent with your current knowledge. Note: monotonic logic still leads to all the "you can't do it" stuff such as the halting theorem, incompleteness, etc. It just prevents mindless backtracking when one tries to use logic as a knowledge base technology. -- Jeff Barnett
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-14 07:28 -0700 |
| Message-ID | <49b0f939-200a-4816-a887-4c39cdbc7cddn@googlegroups.com> |
| In reply to | #35953 |
On Thursday, July 8, 2021 at 4:21:15 PM UTC-4, Jeff Barnett wrote: > On 7/8/2021 6:37 AM, DV wrote: > > On Thursday, July 8, 2021 at 5:44:07 AM UTC-4, pehou...@gmail.com wrote: > >> On Wednesday, July 7, 2021 at 9:19:10 PM UTC-4, Daniel Pehoushek wrote: > >>> simple is good > >>> lifting up symbols halfway above the line is bad > >>> > >>> daniel (little d) > >> monotone reason is linear. > >> abortion is bad. > >> weapons are bad. > > > > Do you mind if I ask what monotone reason is? > A "monotone reasoner" is a system/agent that can learn more but cannot > learn different. Example: It is known that "Bob has a car." We can learn > that "Bob's car is red." but wee cannot deal with "Bob actually has a > skate board, not a car". > > The issue with non-monotonic reasoning is that inferences made in the > past can be invalidated by new knowledge and it's computationally > expensive and god-awful hard to back out those now-wrong results. > > Learning and proving mathematics is in principal monotonic. You can > always rely on old theorems and conclusions as long as you are not > taught something that is inconsistent with your current knowledge. Note: > monotonic logic still leads to all the "you can't do it" stuff such as > the halting theorem, incompleteness, etc. It just prevents mindless > backtracking when one tries to use logic as a knowledge base technology. > -- > Jeff Barnett that is very good jeff. my program bob takes a boolean p formula as input and when the number of satisfying assignments is sufficiently small bob produces a q formula for the p formula which is just precisely a conjunction of all and only universal truths implied by the p formula. tis a giant leap for mankind. part of my own montone reason is all lowercase writing prefereably without punctuation. turns out the "shift" bit on letters behaves a little like the "sign bit" in int architectures, where int is known to be bad. daniel
[toc] | [prev] | [next] | [standalone]
| From | DV <xlt.pjw@gmail.com> |
|---|---|
| Date | 2021-07-14 14:28 -0700 |
| Message-ID | <2b784423-5ed2-4b80-a73d-09ec39f23a07n@googlegroups.com> |
| In reply to | #36302 |
On Wednesday, July 14, 2021 at 10:28:09 AM UTC-4, pehou...@gmail.com wrote: > On Thursday, July 8, 2021 at 4:21:15 PM UTC-4, Jeff Barnett wrote: > > On 7/8/2021 6:37 AM, DV wrote: > > > On Thursday, July 8, 2021 at 5:44:07 AM UTC-4, pehou...@gmail.com wrote: > > >> On Wednesday, July 7, 2021 at 9:19:10 PM UTC-4, Daniel Pehoushek wrote: > > >>> simple is good > > >>> lifting up symbols halfway above the line is bad > > >>> > > >>> daniel (little d) > > >> monotone reason is linear. > > >> abortion is bad. > > >> weapons are bad. > > > > > > Do you mind if I ask what monotone reason is? > > A "monotone reasoner" is a system/agent that can learn more but cannot > > learn different. Example: It is known that "Bob has a car." We can learn > > that "Bob's car is red." but wee cannot deal with "Bob actually has a > > skate board, not a car". > > > > The issue with non-monotonic reasoning is that inferences made in the > > past can be invalidated by new knowledge and it's computationally > > expensive and god-awful hard to back out those now-wrong results. > > > > Learning and proving mathematics is in principal monotonic. You can > > always rely on old theorems and conclusions as long as you are not > > taught something that is inconsistent with your current knowledge. Note: > > monotonic logic still leads to all the "you can't do it" stuff such as > > the halting theorem, incompleteness, etc. It just prevents mindless > > backtracking when one tries to use logic as a knowledge base technology. > > -- > > Jeff Barnett > that is very good jeff. my program bob takes a boolean p formula as input and when the number of satisfying assignments is sufficiently small bob produces a q formula for the p formula which is just precisely a conjunction of all and only universal truths implied by the p formula. tis a giant leap for mankind. > part of my own montone reason is all lowercase writing prefereably without punctuation. > turns out the "shift" bit on letters behaves a little like the "sign bit" in int architectures, where int is known to be bad. > daniel I noticed something that alarmed me in your post, which is that something in your second post sounds a lot like something I was thinking about lately. Obviously, I am not accusing you of having stolen my ideas or something...I am merely pointing out that something you said that I would prefer not to specify sounds quite related to an important thought I had. (I hope you don't mind me pointing that out.) -Philip
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-15 12:51 -0700 |
| Message-ID | <feab1830-7177-43ec-a419-9107c411bb18n@googlegroups.com> |
| In reply to | #36315 |
On Wednesday, July 14, 2021 at 5:28:42 PM UTC-4, DV wrote: > On Wednesday, July 14, 2021 at 10:28:09 AM UTC-4, pehou...@gmail.com wrote: > > On Thursday, July 8, 2021 at 4:21:15 PM UTC-4, Jeff Barnett wrote: > > > On 7/8/2021 6:37 AM, DV wrote: > > > > On Thursday, July 8, 2021 at 5:44:07 AM UTC-4, pehou...@gmail.com wrote: > > > >> On Wednesday, July 7, 2021 at 9:19:10 PM UTC-4, Daniel Pehoushek wrote: > > > >>> simple is good > > > >>> lifting up symbols halfway above the line is bad > > > >>> > > > >>> daniel (little d) > > > >> monotone reason is linear. > > > >> abortion is bad. > > > >> weapons are bad. > > > > > > > > Do you mind if I ask what monotone reason is? > > > A "monotone reasoner" is a system/agent that can learn more but cannot > > > learn different. Example: It is known that "Bob has a car." We can learn > > > that "Bob's car is red." but wee cannot deal with "Bob actually has a > > > skate board, not a car". > > > > > > The issue with non-monotonic reasoning is that inferences made in the > > > past can be invalidated by new knowledge and it's computationally > > > expensive and god-awful hard to back out those now-wrong results. > > > > > > Learning and proving mathematics is in principal monotonic. You can > > > always rely on old theorems and conclusions as long as you are not > > > taught something that is inconsistent with your current knowledge. Note: > > > monotonic logic still leads to all the "you can't do it" stuff such as > > > the halting theorem, incompleteness, etc. It just prevents mindless > > > backtracking when one tries to use logic as a knowledge base technology. > > > -- > > > Jeff Barnett > > that is very good jeff. my program bob takes a boolean p formula as input and when the number of satisfying assignments is sufficiently small bob produces a q formula for the p formula which is just precisely a conjunction of all and only universal truths implied by the p formula. tis a giant leap for mankind. > > part of my own montone reason is all lowercase writing prefereably without punctuation. > > turns out the "shift" bit on letters behaves a little like the "sign bit" in int architectures, where int is known to be bad. > > daniel > I noticed something that alarmed me in your post, which is that something in your second post sounds a lot like something I was thinking about lately. Obviously, I am not accusing you of having stolen my ideas or something...I am merely pointing out that something you said that I would prefer not to specify sounds quite related to an important thought I had. (I hope you don't mind me pointing that out.) > > -Philip go deep dv
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-18 03:45 -0700 |
| Message-ID | <5f4ff994-812e-426d-a764-613ca137f3f2n@googlegroups.com> |
| In reply to | #36353 |
revised relativization, page 185 garey and johnson: P=NP for small n, P!=NP for large N. counting is like a language, small numbers versus large. bob solves small reason completely.
[toc] | [prev] | [next] | [standalone]
| From | DV <xlt.pjw@gmail.com> |
|---|---|
| Date | 2021-07-18 06:48 -0700 |
| Message-ID | <19e00db6-ff59-47d8-964b-bdc685b7c02fn@googlegroups.com> |
| In reply to | #36591 |
On Sunday, July 18, 2021 at 6:45:53 AM UTC-4, pehou...@gmail.com wrote: > revised relativization, page 185 garey and johnson: > P=NP for small n, > P!=NP for large N. > > counting is like a language, small numbers versus large. > bob solves small reason completely. I remember Garey and Johnson...that was the first book I ever got (I checked it out at a library) about P vs. NP or complexity. Do you care to explain what you mean by small reason? Is that about chemistry or quantum mechanics?
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-18 14:28 -0700 |
| Message-ID | <04172d72-96b3-4513-8093-0531970ca6ban@googlegroups.com> |
| In reply to | #36602 |
by small i mean ten boolean propositions with any formula by large i mean one thousand or more i suppose i relativized small and large numbers
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-26 04:45 -0700 |
| Message-ID | <c4889dd6-7fff-4355-b739-c2ff40757b40n@googlegroups.com> |
| In reply to | #36613 |
the model count of a boolean formula tells how many satisfying assignments there are. obtaining the correct model count implies that every solution may be processed. approximating the model count is stupid, ignoring the deep value of correctness. the model counting competition 2021 has much approximation, so much so that they ignored bob who gives only right answers. never has a whole community been on such a boondoggle. On Sunday, July 18, 2021 at 5:28:06 PM UTC-4, Daniel Pehoushek wrote: > by small i mean ten boolean propositions with any formula > by large i mean one thousand or more > > i suppose i relativized small and large numbers
[toc] | [prev] | [next] | [standalone]
| From | Ben Bacarisse <ben.usenet@bsb.me.uk> |
|---|---|
| Date | 2021-07-26 14:35 +0100 |
| Message-ID | <87k0ldb3ka.fsf@bsb.me.uk> |
| In reply to | #37067 |
Daniel Pehoushek <pehoushek1@gmail.com> writes: > the model count of a boolean formula tells how many satisfying > assignments there are. obtaining the correct model count implies that > every solution may be processed. > > approximating the model count is stupid, ignoring the deep value of > correctness. Did bob get the counts right to withing the accuracy of the log_10 estimate? -- Ben.
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-28 00:35 -0700 |
| Message-ID | <4ed2a984-a1d2-4ed4-b182-d6e6c42f4262n@googlegroups.com> |
| In reply to | #37072 |
On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: > Daniel Pehoushek <pehou...@gmail.com> writes: > > > the model count of a boolean formula tells how many satisfying > > assignments there are. obtaining the correct model count implies that > > every solution may be processed. > > > > approximating the model count is stupid, ignoring the deep value of > > correctness. > Did bob get the counts right to withing the accuracy of the log_10 > estimate? > > -- > Ben. approximation is just plain wrong. think about this: the correct model count implies access to every bit of truth in every solution. the correct model count has a large implied truth structure with N*#P bits. approximations of #P have zero such truth structures. the correct model count is like a correct debit card number allowing access to money. an approximate count is like an approximate card number, worthless. bob gets the counts right. zero wrong answers. zero approximation. the log10 estimate is stupid. fuzzes the whole answer system. its there to make approximation look better. most of the benchmarks are impossible to compute in allotted time. the few that are solvable then get fuzzed by log10. if i ran the contest the benchmarks would all be more solvable. track1 would be exact correct answers only, one wrong and out. track1 would be the correctness track. track2 would be the incorrectness track. any program, including approximating programs, could be submitted to track 2. no one would care. i am trying to talk with the committee, but they love approximation, which is incorrect at the core. daniel
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-28 02:40 -0700 |
| Message-ID | <01c87d31-992c-4dc5-8a15-d171f9a165b9n@googlegroups.com> |
| In reply to | #37181 |
my latest try to get a response from the committee: i study universal truth. approximation of truth is an oxymoron. for the model counting competition 2021 i propose: track 1: the correctness track track 2: the incorrectness track (approximations) On Wednesday, July 28, 2021 at 3:35:53 AM UTC-4, Daniel Pehoushek wrote: > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: > > Daniel Pehoushek <pehou...@gmail.com> writes: > > > > > the model count of a boolean formula tells how many satisfying > > > assignments there are. obtaining the correct model count implies that > > > every solution may be processed. > > > > > > approximating the model count is stupid, ignoring the deep value of > > > correctness. > > Did bob get the counts right to withing the accuracy of the log_10 > > estimate? > > > > -- > > Ben. > approximation is just plain wrong. think about this: the correct model count > implies access to every bit of truth in every solution. the correct > model count has a large implied truth structure with N*#P bits. > approximations of #P have zero such truth structures. the correct model > count is like a correct debit card number allowing access to money. > an approximate count is like an approximate card number, worthless. > > bob gets the counts right. zero wrong answers. zero approximation. > the log10 estimate is stupid. fuzzes the whole answer system. > its there to make approximation look better. > > most of the benchmarks are impossible to compute in allotted time. > the few that are solvable then get fuzzed by log10. > > if i ran the contest the benchmarks would all be more solvable. > track1 would be exact correct answers only, one wrong and out. > > track1 would be the correctness track. > track2 would be the incorrectness track. > > any program, including approximating programs, could be > submitted to track 2. no one would care. > > i am trying to talk with the committee, but they love > approximation, which is incorrect at the core. > daniel
[toc] | [prev] | [next] | [standalone]
| From | Ben Bacarisse <ben.usenet@bsb.me.uk> |
|---|---|
| Date | 2021-07-28 17:27 +0100 |
| Message-ID | <87bl6m8ktw.fsf@bsb.me.uk> |
| In reply to | #37181 |
Daniel Pehoushek <pehoushek1@gmail.com> writes: > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: >> Daniel Pehoushek <pehou...@gmail.com> writes: >> >> > the model count of a boolean formula tells how many satisfying >> > assignments there are. obtaining the correct model count implies that >> > every solution may be processed. >> > >> > approximating the model count is stupid, ignoring the deep value of >> > correctness. >> >> Did bob get the counts right to withing the accuracy of the log_10 >> estimate? > > approximation is just plain wrong. Just a heads-up: you are more likely to get people engaging with what you say if you answer their questions. > bob gets the counts right. So is that a "yes"? Did bob's counts match the figures expected from competition entrants or not? -- Ben.
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-28 11:50 -0700 |
| Message-ID | <8be0b3ce-0a82-4440-b03c-b1873ecc4cbfn@googlegroups.com> |
| In reply to | #37193 |
On Wednesday, July 28, 2021 at 12:27:57 PM UTC-4, Ben Bacarisse wrote: > Daniel Pehoushek <pehou...@gmail.com> writes: > > > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: > >> Daniel Pehoushek <pehou...@gmail.com> writes: > >> > >> > the model count of a boolean formula tells how many satisfying > >> > assignments there are. obtaining the correct model count implies that > >> > every solution may be processed. > >> > > >> > approximating the model count is stupid, ignoring the deep value of > >> > correctness. > >> > >> Did bob get the counts right to withing the accuracy of the log_10 > >> estimate? > > > > approximation is just plain wrong. > Just a heads-up: you are more likely to get people engaging with what > you say if you answer their questions. > > bob gets the counts right. > So is that a "yes"? Did bob's counts match the figures expected from > competition entrants or not? there were not very many benchmarks solvable in the allotted time by an exact solver like bob, but bob got all those right. the problem is for some reason they love approximate answers. so programs that answered everything, but only approximately, are doing well. the committee has never given me any competition results report. i have been trying to get any reply from them, but nothing so far. they don't tell me how well bob did; something is wrong, because all i get are null responses. daniel > > -- > Ben.
[toc] | [prev] | [next] | [standalone]
| From | Ben Bacarisse <ben.usenet@bsb.me.uk> |
|---|---|
| Date | 2021-07-28 20:52 +0100 |
| Message-ID | <875ywu8bdq.fsf@bsb.me.uk> |
| In reply to | #37209 |
Daniel Pehoushek <pehoushek1@gmail.com> writes: > On Wednesday, July 28, 2021 at 12:27:57 PM UTC-4, Ben Bacarisse wrote: >> Daniel Pehoushek <pehou...@gmail.com> writes: >> >> > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: >> >> Daniel Pehoushek <pehou...@gmail.com> writes: >> >> >> >> > the model count of a boolean formula tells how many satisfying >> >> > assignments there are. obtaining the correct model count implies that >> >> > every solution may be processed. >> >> > >> >> > approximating the model count is stupid, ignoring the deep value of >> >> > correctness. >> >> >> >> Did bob get the counts right to withing the accuracy of the log_10 >> >> estimate? >> > >> > approximation is just plain wrong. >> Just a heads-up: you are more likely to get people engaging with what >> you say if you answer their questions. >> > bob gets the counts right. >> So is that a "yes"? Did bob's counts match the figures expected from >> competition entrants or not? > > there were not very many benchmarks solvable in the allotted time > by an exact solver like bob, but bob got all those right. > > the problem is for some reason they love approximate answers. > so programs that answered everything, but only approximately, > are doing well. Maybe this was not the right competition for bob. What did the requirements say a solver had to do? > the committee has never given me any competition results report. > i have been trying to get any reply from them, but nothing so far. > they don't tell me how well bob did; something is wrong, > because all i get are null responses. Do you have someone who can review your submissions to the committee? If they are anything like your posting here, the committee may not have clue about what you are saying. -- Ben.
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-28 15:53 -0700 |
| Message-ID | <db8b0dc8-1657-454d-a48f-fe4dc4c267c4n@googlegroups.com> |
| In reply to | #37211 |
the name is model counting competition 2021. bob counts models. solving the #P problem. most of the entries only approximate, because real counting is too hard. the committee accepted approximations as readily as real model counts. that is where i have been arguing with them. i claim real model counts are more valuable... but its like pulling teeth. the committee likes to have answers, so the more answers a program gives, even if the answers are approximate, the more the committee likes them. if i ran things, there would be two tracks: track 1: correctness only track (bob would win) track 2: incorrectness track (for approximations) to me and bob, approximations are complete bullshit, with zero relation to the real implied truth structure of real models. daniel On Wednesday, July 28, 2021 at 3:52:03 PM UTC-4, Ben Bacarisse wrote: > Daniel Pehoushek <pehou...@gmail.com> writes: > > > On Wednesday, July 28, 2021 at 12:27:57 PM UTC-4, Ben Bacarisse wrote: > >> Daniel Pehoushek <pehou...@gmail.com> writes: > >> > >> > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: > >> >> Daniel Pehoushek <pehou...@gmail.com> writes: > >> >> > >> >> > the model count of a boolean formula tells how many satisfying > >> >> > assignments there are. obtaining the correct model count implies that > >> >> > every solution may be processed. > >> >> > > >> >> > approximating the model count is stupid, ignoring the deep value of > >> >> > correctness. > >> >> > >> >> Did bob get the counts right to withing the accuracy of the log_10 > >> >> estimate? > >> > > >> > approximation is just plain wrong. > >> Just a heads-up: you are more likely to get people engaging with what > >> you say if you answer their questions. > >> > bob gets the counts right. > >> So is that a "yes"? Did bob's counts match the figures expected from > >> competition entrants or not? > > > > there were not very many benchmarks solvable in the allotted time > > by an exact solver like bob, but bob got all those right. > > > > the problem is for some reason they love approximate answers. > > so programs that answered everything, but only approximately, > > are doing well. > Maybe this was not the right competition for bob. What did the > requirements say a solver had to do? > > the committee has never given me any competition results report. > > i have been trying to get any reply from them, but nothing so far. > > they don't tell me how well bob did; something is wrong, > > because all i get are null responses. > Do you have someone who can review your submissions to the committee? > If they are anything like your posting here, the committee may not have > clue about what you are saying. > > -- > Ben.
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-28 16:58 -0700 |
| Message-ID | <17d77cd8-cbc8-4932-be1d-f909c3297c6an@googlegroups.com> |
| In reply to | #37226 |
the problem is that in the track 1 correctness track they counted approximate random guesses as more valuable than "no answer yet"in allotted time. On Wednesday, July 28, 2021 at 6:53:10 PM UTC-4, Daniel Pehoushek wrote: > the name is model counting competition 2021. > bob counts models. solving the #P problem. > most of the entries only approximate, > because real counting is too hard. > > the committee accepted approximations as readily as real model counts. > that is where i have been arguing with them. > i claim real model counts are more valuable... > but its like pulling teeth. the committee likes > to have answers, so the more answers a program gives, > even if the answers are approximate, the more the > committee likes them. > > if i ran things, there would be two tracks: > track 1: correctness only track (bob would win) > track 2: incorrectness track (for approximations) > > to me and bob, approximations are complete bullshit, with zero relation > to the real implied truth structure of real models. > daniel > On Wednesday, July 28, 2021 at 3:52:03 PM UTC-4, Ben Bacarisse wrote: > > Daniel Pehoushek <pehou...@gmail.com> writes: > > > > > On Wednesday, July 28, 2021 at 12:27:57 PM UTC-4, Ben Bacarisse wrote: > > >> Daniel Pehoushek <pehou...@gmail.com> writes: > > >> > > >> > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: > > >> >> Daniel Pehoushek <pehou...@gmail.com> writes: > > >> >> > > >> >> > the model count of a boolean formula tells how many satisfying > > >> >> > assignments there are. obtaining the correct model count implies that > > >> >> > every solution may be processed. > > >> >> > > > >> >> > approximating the model count is stupid, ignoring the deep value of > > >> >> > correctness. > > >> >> > > >> >> Did bob get the counts right to withing the accuracy of the log_10 > > >> >> estimate? > > >> > > > >> > approximation is just plain wrong. > > >> Just a heads-up: you are more likely to get people engaging with what > > >> you say if you answer their questions. > > >> > bob gets the counts right. > > >> So is that a "yes"? Did bob's counts match the figures expected from > > >> competition entrants or not? > > > > > > there were not very many benchmarks solvable in the allotted time > > > by an exact solver like bob, but bob got all those right. > > > > > > the problem is for some reason they love approximate answers. > > > so programs that answered everything, but only approximately, > > > are doing well. > > Maybe this was not the right competition for bob. What did the > > requirements say a solver had to do? > > > the committee has never given me any competition results report. > > > i have been trying to get any reply from them, but nothing so far. > > > they don't tell me how well bob did; something is wrong, > > > because all i get are null responses. > > Do you have someone who can review your submissions to the committee? > > If they are anything like your posting here, the committee may not have > > clue about what you are saying. > > > > -- > > Ben.
[toc] | [prev] | [next] | [standalone]
| From | Daniel Pehoushek <pehoushek1@gmail.com> |
|---|---|
| Date | 2021-07-28 17:41 -0700 |
| Message-ID | <bcf74a9c-7ad9-4231-b906-4131c8cd2ce4n@googlegroups.com> |
| In reply to | #37233 |
ben, is this clear? in track 1 the correctness track approximation is senseless is what i am saying in track 2 the incorrectness track approximation is fine On Wednesday, July 28, 2021 at 7:58:47 PM UTC-4, Daniel Pehoushek wrote: > the problem is that in the track 1 correctness track > they counted approximate random guesses > as more valuable than "no answer yet"in allotted time. > On Wednesday, July 28, 2021 at 6:53:10 PM UTC-4, Daniel Pehoushek wrote: > > the name is model counting competition 2021. > > bob counts models. solving the #P problem. > > most of the entries only approximate, > > because real counting is too hard. > > > > the committee accepted approximations as readily as real model counts. > > that is where i have been arguing with them. > > i claim real model counts are more valuable... > > but its like pulling teeth. the committee likes > > to have answers, so the more answers a program gives, > > even if the answers are approximate, the more the > > committee likes them. > > > > if i ran things, there would be two tracks: > > track 1: correctness only track (bob would win) > > track 2: incorrectness track (for approximations) > > > > to me and bob, approximations are complete bullshit, with zero relation > > to the real implied truth structure of real models. > > daniel > > On Wednesday, July 28, 2021 at 3:52:03 PM UTC-4, Ben Bacarisse wrote: > > > Daniel Pehoushek <pehou...@gmail.com> writes: > > > > > > > On Wednesday, July 28, 2021 at 12:27:57 PM UTC-4, Ben Bacarisse wrote: > > > >> Daniel Pehoushek <pehou...@gmail.com> writes: > > > >> > > > >> > On Monday, July 26, 2021 at 9:35:51 AM UTC-4, Ben Bacarisse wrote: > > > >> >> Daniel Pehoushek <pehou...@gmail.com> writes: > > > >> >> > > > >> >> > the model count of a boolean formula tells how many satisfying > > > >> >> > assignments there are. obtaining the correct model count implies that > > > >> >> > every solution may be processed. > > > >> >> > > > > >> >> > approximating the model count is stupid, ignoring the deep value of > > > >> >> > correctness. > > > >> >> > > > >> >> Did bob get the counts right to withing the accuracy of the log_10 > > > >> >> estimate? > > > >> > > > > >> > approximation is just plain wrong. > > > >> Just a heads-up: you are more likely to get people engaging with what > > > >> you say if you answer their questions. > > > >> > bob gets the counts right. > > > >> So is that a "yes"? Did bob's counts match the figures expected from > > > >> competition entrants or not? > > > > > > > > there were not very many benchmarks solvable in the allotted time > > > > by an exact solver like bob, but bob got all those right. > > > > > > > > the problem is for some reason they love approximate answers. > > > > so programs that answered everything, but only approximately, > > > > are doing well. > > > Maybe this was not the right competition for bob. What did the > > > requirements say a solver had to do? > > > > the committee has never given me any competition results report. > > > > i have been trying to get any reply from them, but nothing so far. > > > > they don't tell me how well bob did; something is wrong, > > > > because all i get are null responses. > > > Do you have someone who can review your submissions to the committee? > > > If they are anything like your posting here, the committee may not have > > > clue about what you are saying. > > > > > > -- > > > Ben.
[toc] | [prev] | [next] | [standalone]
Page 1 of 3 [1] 2 3 Next page →
Back to top | Article view | comp.theory
csiph-web