marsultor/ee-work
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

this is getting tedious

Browse Source
This commit is contained in:
Mars Ultor
2025-02-28 03:01:12 -06:00
parent 8a21ebb22f
commit 983b76b029
11 changed files with 304 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
EEMLA.aux
View File

@@ -2,6 +2,19 @@
\abx@aux@refcontext{nty/global//global/global}
\abx@aux@cite{0}{app14020744}
\abx@aux@segm{0}{0}{app14020744}
\abx@aux@read@bbl@mdfivesum{AEE4F911165DE78F7C9D7634D26F9914}
\abx@aux@cite{0}{Wang2024}
\abx@aux@segm{0}{0}{Wang2024}
\abx@aux@cite{0}{vaswani2023attentionneed}
\abx@aux@segm{0}{0}{vaswani2023attentionneed}
\abx@aux@cite{0}{Wang2024}
\abx@aux@segm{0}{0}{Wang2024}
\abx@aux@cite{0}{ivanov2024}
\abx@aux@segm{0}{0}{ivanov2024}
\abx@aux@cite{0}{ivanov2024}
\abx@aux@segm{0}{0}{ivanov2024}
\abx@aux@read@bbl@mdfivesum{A739BAAE76801A6EB23A4AE3E4219B4A}
\abx@aux@defaultrefcontext{0}{ivanov2024}{nty/global//global/global}
\abx@aux@defaultrefcontext{0}{app14020744}{nty/global//global/global}
\gdef \@abspage@last{2}
\abx@aux@defaultrefcontext{0}{vaswani2023attentionneed}{nty/global//global/global}
\abx@aux@defaultrefcontext{0}{Wang2024}{nty/global//global/global}
\gdef \@abspage@last{5}

174
EEMLA.bbl
View File

@@ -19,6 +19,46 @@
\refsection{0}
\datalist[entry]{nty/global//global/global}
\entry{ivanov2024}{misc}{}
\name{author}{2}{}{%
{{un=0,uniquepart=base,hash=1935b6f0043d4bac823842ff5d478faf}{%
family={Ivanov},
familyi={I\bibinitperiod},
given={Todor},
giveni={T\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=495d2036e580863cb79ad80a7bb2fc78}{%
family={Penchev},
familyi={P\bibinitperiod},
given={Valeri},
giveni={V\bibinitperiod},
givenun=0}}%
}
\strng{namehash}{259e51507bbcab245b7267c088c1998f}
\strng{fullhash}{259e51507bbcab245b7267c088c1998f}
\strng{bibnamehash}{259e51507bbcab245b7267c088c1998f}
\strng{authorbibnamehash}{259e51507bbcab245b7267c088c1998f}
\strng{authornamehash}{259e51507bbcab245b7267c088c1998f}
\strng{authorfullhash}{259e51507bbcab245b7267c088c1998f}
\field{sortinit}{I}
\field{sortinithash}{8d291c51ee89b6cd86bf5379f0b151d8}
\true{singletitle}
\field{labelnamesource}{author}
\field{labeltitlesource}{title}
\field{eprintclass}{cs.DC}
\field{eprinttype}{arXiv}
\field{title}{AI Benchmarks and Datasets for LLM Evaluation}
\field{year}{2024}
\verb{eprint}
\verb 2412.01020
\endverb
\verb{urlraw}
\verb arxiv.org/abs/2412.01020
\endverb
\verb{url}
\verb arxiv.org/abs/2412.01020
\endverb
\endentry
\entry{app14020744}{article}{}
\name{author}{1}{}{%
{{un=0,uniquepart=base,hash=d8c43e5429158fe51408ffa847a4a856}{%
@@ -50,6 +90,140 @@
\verb 10.3390/app14020744
\endverb
\endentry
\entry{vaswani2023attentionneed}{misc}{}
\name{author}{8}{}{%
{{un=0,uniquepart=base,hash=7f28e84700536646dd6620a0db07ad09}{%
family={Vaswani},
familyi={V\bibinitperiod},
given={Ashish},
giveni={A\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=62efade83d70f0323fe248755e6c90c5}{%
family={Shazeer},
familyi={S\bibinitperiod},
given={Noam},
giveni={N\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=06649ebab1ea5cac0250746a19764975}{%
family={Parmar},
familyi={P\bibinitperiod},
given={Niki},
giveni={N\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=831027ee0ebf22375e2a86afc1881909}{%
family={Uszkoreit},
familyi={U\bibinitperiod},
given={Jakob},
giveni={J\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=2fd2982e30ebcec93ec1cf76e0d797fd}{%
family={Jones},
familyi={J\bibinitperiod},
given={Llion},
giveni={L\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=27b07e4eacbf4ef7a1438e3badb7dd8d}{%
family={Gomez},
familyi={G\bibinitperiod},
given={Aidan\bibnamedelima N.},
giveni={A\bibinitperiod\bibinitdelim N\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=f2bc899b1160163417da7bf510f15d33}{%
family={Kaiser},
familyi={K\bibinitperiod},
given={Lukasz},
giveni={L\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=95595a0fefb86187cbc36e551017d332}{%
family={Polosukhin},
familyi={P\bibinitperiod},
given={Illia},
giveni={I\bibinitperiod},
givenun=0}}%
}
\strng{namehash}{ee273ab30cfb889666f8c4d806eb9ce7}
\strng{fullhash}{f82970bbd2bdd7a002d2af62b743d5cc}
\strng{bibnamehash}{ee273ab30cfb889666f8c4d806eb9ce7}
\strng{authorbibnamehash}{ee273ab30cfb889666f8c4d806eb9ce7}
\strng{authornamehash}{ee273ab30cfb889666f8c4d806eb9ce7}
\strng{authorfullhash}{f82970bbd2bdd7a002d2af62b743d5cc}
\field{sortinit}{V}
\field{sortinithash}{afb52128e5b4dc4b843768c0113d673b}
\true{singletitle}
\field{labelnamesource}{author}
\field{labeltitlesource}{title}
\field{eprintclass}{cs.CL}
\field{eprinttype}{arXiv}
\field{title}{Attention Is All You Need}
\field{year}{2023}
\verb{eprint}
\verb 1706.03762
\endverb
\verb{urlraw}
\verb arxiv.org/abs/1706.03762
\endverb
\verb{url}
\verb arxiv.org/abs/1706.03762
\endverb
\endentry
\entry{Wang2024}{article}{}
\name{author}{6}{}{%
{{un=0,uniquepart=base,hash=7cca10cee48e9c197439e4af610acfe5}{%
family={Wang},
familyi={W\bibinitperiod},
given={Zichong},
giveni={Z\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=bc41c1f771e246b6d8ce744badb182e3}{%
family={Chu},
familyi={C\bibinitperiod},
given={Zhibo},
giveni={Z\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=82b0035db67db8bd400c34e8a5eec07b}{%
family={Doan},
familyi={D\bibinitperiod},
given={Thang\bibnamedelima Viet},
giveni={T\bibinitperiod\bibinitdelim V\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=852b650254c75a15c1024df13b29189c}{%
family={Ni},
familyi={N\bibinitperiod},
given={Shiwen},
giveni={S\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=feb96ca112c179e320db2db693e022b8}{%
family={Yang},
familyi={Y\bibinitperiod},
given={Min},
giveni={M\bibinitperiod},
givenun=0}}%
{{un=0,uniquepart=base,hash=dd4baede28b306ab6d37dd79d89a935b}{%
family={Zhang},
familyi={Z\bibinitperiod},
given={Wenbin},
giveni={W\bibinitperiod},
givenun=0}}%
}
\strng{namehash}{1c3e58e991d8f7a6ae5aee1e95c5cd8a}
\strng{fullhash}{391aec39c1c26e8d5e7517c1ab227456}
\strng{bibnamehash}{1c3e58e991d8f7a6ae5aee1e95c5cd8a}
\strng{authorbibnamehash}{1c3e58e991d8f7a6ae5aee1e95c5cd8a}
\strng{authornamehash}{1c3e58e991d8f7a6ae5aee1e95c5cd8a}
\strng{authorfullhash}{391aec39c1c26e8d5e7517c1ab227456}
\field{sortinit}{W}
\field{sortinithash}{4315d78024d0cea9b57a0c6f0e35ed0d}
\true{singletitle}
\field{labelnamesource}{author}
\field{labeltitlesource}{title}
\field{journaltitle}{AI and Ethics}
\field{month}{10}
\field{title}{History, development, and principles of large language models: An introductory survey}
\field{year}{2024}
\verb{doi}
\verb 10.1007/s43681-024-00583-7
\endverb
\endentry
\enddatalist
\endrefsection
\endinput

5
EEMLA.bcf
View File

@@ -2765,6 +2765,11 @@
</bcf:bibdata>
<bcf:section number="0">
<bcf:citekey order="1" intorder="1">app14020744</bcf:citekey>
<bcf:citekey order="2" intorder="1">Wang2024</bcf:citekey>
<bcf:citekey order="3" intorder="1">vaswani2023attentionneed</bcf:citekey>
<bcf:citekey order="4" intorder="1">Wang2024</bcf:citekey>
<bcf:citekey order="5" intorder="1">ivanov2024</bcf:citekey>
<bcf:citekey order="6" intorder="1">ivanov2024</bcf:citekey>
</bcf:section>
<!-- SORTING TEMPLATES -->
<bcf:sortingtemplate name="nty">

28
EEMLA.blg
View File

@@ -1,15 +1,17 @@
[0] Config.pm:307> INFO - This is Biber 2.19
[0] Config.pm:310> INFO - Logfile is 'EEMLA.blg'
[37] biber:340> INFO - === Fri Feb 28, 2025, 01:10:48
[45] Biber.pm:419> INFO - Reading 'EEMLA.bcf'
[79] Biber.pm:979> INFO - Found 1 citekeys in bib section 0
[86] Biber.pm:4419> INFO - Processing section 0
[93] Biber.pm:4610> INFO - Looking for bibtex file 'references.bib' for section 0
[93] bibtex.pm:1713> INFO - LaTeX decoding ...
[94] bibtex.pm:1519> INFO - Found BibTeX data source 'references.bib'
[112] UCollate.pm:68> INFO - Overriding locale 'en-US' defaults 'variable = shifted' with 'variable = non-ignorable'
[112] UCollate.pm:68> INFO - Overriding locale 'en-US' defaults 'normalization = NFD' with 'normalization = prenormalized'
[112] Biber.pm:4239> INFO - Sorting list 'nty/global//global/global' of type 'entry' with template 'nty' and locale 'en-US'
[112] Biber.pm:4245> INFO - No sort tailoring available for locale 'en-US'
[113] bbl.pm:660> INFO - Writing 'EEMLA.bbl' with encoding 'UTF-8'
[113] bbl.pm:763> INFO - Output to EEMLA.bbl
[41] biber:340> INFO - === Fri Feb 28, 2025, 02:56:54
[50] Biber.pm:419> INFO - Reading 'EEMLA.bcf'
[86] Biber.pm:979> INFO - Found 4 citekeys in bib section 0
[95] Biber.pm:4419> INFO - Processing section 0
[100] Biber.pm:4610> INFO - Looking for bibtex file 'references.bib' for section 0
[100] bibtex.pm:1713> INFO - LaTeX decoding ...
[103] bibtex.pm:1519> INFO - Found BibTeX data source 'references.bib'
[137] UCollate.pm:68> INFO - Overriding locale 'en-US' defaults 'variable = shifted' with 'variable = non-ignorable'
[137] UCollate.pm:68> INFO - Overriding locale 'en-US' defaults 'normalization = NFD' with 'normalization = prenormalized'
[137] Biber.pm:4239> INFO - Sorting list 'nty/global//global/global' of type 'entry' with template 'nty' and locale 'en-US'
[137] Biber.pm:4245> INFO - No sort tailoring available for locale 'en-US'
[142] bbl.pm:660> INFO - Writing 'EEMLA.bbl' with encoding 'UTF-8'
[143] bbl.pm:763> INFO - Output to EEMLA.bbl
[143] Biber.pm:131> WARN - legacy month field 'Oct' in entry 'Wang2024' is not an integer - this will probably not sort properly.
[143] Biber.pm:133> INFO - WARNINGS: 1

21
EEMLA.fdb_latexmk
View File

@@ -1,12 +1,12 @@
# Fdb version 4
["biber EEMLA"] 1740726648.62462 "EEMLA.bcf" "EEMLA.bbl" "EEMLA" 1740727159.44349 0
"EEMLA.bcf" 1740727159.40986 125151 7d19eab9f557d086de927c33fead835b "pdflatex"
"references.bib" 1740725712.24895 1400 e27bc3a6c0b47f6ed6d0043dd6ccb05e ""
["biber EEMLA"] 1740733013.7148 "EEMLA.bcf" "EEMLA.bbl" "EEMLA" 1740733218.14824 0
"EEMLA.bcf" 1740733218.11349 125486 fd2cc3ea398de230c8b44dae74685005 "pdflatex"
"references.bib" 1740732945.89131 2369 a24b87b373146d16ecacc3c3e385073c ""
(generated)
"EEMLA.bbl"
"EEMLA.blg"
(rewritten before read)
["pdflatex"] 1740727159.00655 "EEMLA.tex" "EEMLA.pdf" "EEMLA" 1740727159.44373 0
["pdflatex"] 1740733217.68731 "EEMLA.tex" "EEMLA.pdf" "EEMLA" 1740733218.14847 0
"/usr/share/texmf-dist/fonts/enc/dvips/base/8r.enc" 1736268207 4850 80dc9bab7f31fb78a000ccfed0e27cab ""
"/usr/share/texmf-dist/fonts/map/fontname/texfonts.map" 1736268207 3524 cb3e574dea2d1052e39280babc910dc8 ""
"/usr/share/texmf-dist/fonts/tfm/adobe/times/ptmb7t.tfm" 1736268207 2172 fd0c924230362ff848a33632ed45dc23 ""
@@ -15,6 +15,9 @@
"/usr/share/texmf-dist/fonts/tfm/adobe/times/ptmr8r.tfm" 1736268207 4408 25b74d011a4c66b7f212c0cc3c90061b ""
"/usr/share/texmf-dist/fonts/tfm/adobe/times/ptmri7t.tfm" 1736268207 2288 f478fc8fed18759effb59f3dad7f3084 ""
"/usr/share/texmf-dist/fonts/tfm/adobe/times/ptmri8r.tfm" 1736268207 4640 532ca3305aad10cc01d769f3f91f1029 ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmbx12.tfm" 1736268207 1324 c910af8c371558dc20f2d7822f66fe64 ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmbx6.tfm" 1736268207 1344 8a0be4fe4d376203000810ad4dc81558 ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmbx8.tfm" 1736268207 1332 1fde11373e221473104d6cc5993f046e ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmmi12.tfm" 1736268207 1524 4414a8315f39513458b80dfc63bff03a ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmmi6.tfm" 1736268207 1512 f21f83efb36853c0b70002322c1ab3ad ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmmi8.tfm" 1736268207 1520 eccf95517727cb11801f4f1aee3a21b4 ""
@@ -24,11 +27,13 @@
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmsy10.tfm" 1736268207 1124 6c73e740cf17375f03eec0ee63599741 ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmsy6.tfm" 1736268207 1116 933a60c408fc0a863a92debe84b2d294 ""
"/usr/share/texmf-dist/fonts/tfm/public/cm/cmsy8.tfm" 1736268207 1120 8b7d695260f3cff42e636090a8002094 ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmbx12.pfb" 1736268207 32080 340ef9bf63678554ee606688e7b5339d ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmex10.pfb" 1736268207 30251 6afa5cb1d0204815a708a080681d4674 ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi12.pfb" 1736268207 36741 fa121aac0049305630cf160b86157ee4 ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi8.pfb" 1736268207 35469 70d41d2b9ea31d5d813066df7c99281c ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr12.pfb" 1736268207 32722 d7379af29a190c3f453aba36302ff5a9 ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr8.pfb" 1736268207 32726 0a1aea6fcd6468ee2cf64d891f5c43c8 ""
"/usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmsy10.pfb" 1736268207 32569 5e5ddc8df908dea60932f3c484a54c0d ""
"/usr/share/texmf-dist/fonts/type1/urw/times/utmb8a.pfb" 1736268207 44729 811d6c62865936705a31c797a1d5dada ""
"/usr/share/texmf-dist/fonts/type1/urw/times/utmr8a.pfb" 1736268207 46026 6dab18b61c907687b520c72847215a68 ""
"/usr/share/texmf-dist/fonts/type1/urw/times/utmri8a.pfb" 1736268207 45458 a3faba884469519614ca56ba5f6b1de1 ""
@@ -84,10 +89,10 @@
"/usr/share/texmf-dist/web2c/texmf.cnf" 1736268207 41588 b43d3e860a4f94167ee1e725ff526a72 ""
"/var/lib/texmf/fonts/map/pdftex/updmap/pdftex.map" 1740110826.09194 5318467 8236939515598d005dfbcc2e1a40f206 ""
"/var/lib/texmf/web2c/pdftex/pdflatex.fmt" 1740110801 7112991 1dc785529b0f131eb9fafe88e40ba344 ""
"EEMLA.aux" 1740727159.40946 274 9c84998d6d05c7c52d04978f0fca6c6c "pdflatex"
"EEMLA.bbl" 1740726649.04747 2804 aee4f911165de78f7c9d7634d26f9914 "biber EEMLA"
"EEMLA.run.xml" 1740727159.411 2437 a20ebb97c6b6df9ec77b4a047cab8b74 "pdflatex"
"EEMLA.tex" 1740727158.48416 2806 49474dbb1c44067a2c1524217ff7564f ""
"EEMLA.aux" 1740733218.11256 818 607726c98a04d660d760eb9790b6e498 "pdflatex"
"EEMLA.bbl" 1740733014.15262 9373 a739baae76801a6eb23a4ae3e4219b4a "biber EEMLA"
"EEMLA.run.xml" 1740733218.11417 2437 a20ebb97c6b6df9ec77b4a047cab8b74 "pdflatex"
"EEMLA.tex" 1740733217.00648 9156 7874abaeef9e25ad97e500f377e8f691 ""
(generated)
"EEMLA.aux"
"EEMLA.bcf"

5
EEMLA.fls
View File

@@ -139,6 +139,9 @@ INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmmi6.tfm
INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmsy10.tfm
INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmsy8.tfm
INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmsy6.tfm
INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmbx12.tfm
INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmbx8.tfm
INPUT /usr/share/texmf-dist/fonts/tfm/public/cm/cmbx6.tfm
OUTPUT EEMLA.pdf
INPUT /usr/share/texmf-dist/fonts/vf/adobe/times/ptmr7t.vf
INPUT /usr/share/texmf-dist/fonts/tfm/adobe/times/ptmr8r.tfm
@@ -152,11 +155,13 @@ INPUT /usr/share/texmf-dist/fonts/tfm/adobe/times/ptmri8r.tfm
INPUT EEMLA.aux
INPUT EEMLA.run.xml
OUTPUT EEMLA.run.xml
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmbx12.pfb
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmex10.pfb
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi12.pfb
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi8.pfb
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr12.pfb
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr8.pfb
INPUT /usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmsy10.pfb
INPUT /usr/share/texmf-dist/fonts/type1/urw/times/utmb8a.pfb
INPUT /usr/share/texmf-dist/fonts/type1/urw/times/utmr8a.pfb
INPUT /usr/share/texmf-dist/fonts/type1/urw/times/utmri8a.pfb

48
EEMLA.log
View File

@@ -1,4 +1,4 @@
This is pdfTeX, Version 3.141592653-2.6-1.40.26 (TeX Live 2024/Arch Linux) (preloaded format=pdflatex 2025.2.20) 28 FEB 2025 01:19
This is pdfTeX, Version 3.141592653-2.6-1.40.26 (TeX Live 2024/Arch Linux) (preloaded format=pdflatex 2025.2.20) 28 FEB 2025 03:00
entering extended mode
restricted \write18 enabled.
file:line:error style messages enabled.
@@ -475,7 +475,31 @@ Package fancyhdr Warning: \headheight is too small (12.0pt):
(fancyhdr) You might also make \topmargin smaller to compensate:
(fancyhdr) \addtolength{\topmargin}{-2.49998pt}.
[2] (./EEMLA.aux)
[2]
Package fancyhdr Warning: \headheight is too small (12.0pt):
(fancyhdr) Make it at least 14.49998pt, for example:
(fancyhdr) \setlength{\headheight}{14.49998pt}.
(fancyhdr) You might also make \topmargin smaller to compensate:
(fancyhdr) \addtolength{\topmargin}{-2.49998pt}.
[3]
Package fancyhdr Warning: \headheight is too small (12.0pt):
(fancyhdr) Make it at least 14.49998pt, for example:
(fancyhdr) \setlength{\headheight}{14.49998pt}.
(fancyhdr) You might also make \topmargin smaller to compensate:
(fancyhdr) \addtolength{\topmargin}{-2.49998pt}.
[4]
Package fancyhdr Warning: \headheight is too small (12.0pt):
(fancyhdr) Make it at least 14.49998pt, for example:
(fancyhdr) \setlength{\headheight}{14.49998pt}.
(fancyhdr) You might also make \topmargin smaller to compensate:
(fancyhdr) \addtolength{\topmargin}{-2.49998pt}.
[5] (./EEMLA.aux)
***********
LaTeX2e <2023-11-01> patch level 1
L3 programming layer <2024-02-20>
@@ -485,18 +509,18 @@ Package logreq Info: Writing requests to 'EEMLA.run.xml'.
)
Here is how much of TeX's memory you used:
10781 strings out of 476076
204959 string characters out of 5793775
1931187 words of memory out of 5000000
32796 multiletter control sequences out of 15000+600000
565700 words of font info for 51 fonts, out of 8000000 for 9000
10810 strings out of 476076
205922 string characters out of 5793775
1933187 words of memory out of 5000000
32822 multiletter control sequences out of 15000+600000
566628 words of font info for 54 fonts, out of 8000000 for 9000
14 hyphenation exceptions out of 8191
72i,11n,81p,1102b,2647s stack positions out of 10000i,1000n,20000p,200000b,200000s
</usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmex10.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi12.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi8.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr12.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr8.pfb></usr/share/texmf-dist/fonts/type1/urw/times/utmb8a.pfb></usr/share/texmf-dist/fonts/type1/urw/times/utmr8a.pfb></usr/share/texmf-dist/fonts/type1/urw/times/utmri8a.pfb>
Output written on EEMLA.pdf (2 pages, 81151 bytes).
72i,11n,81p,1163b,2647s stack positions out of 10000i,1000n,20000p,200000b,200000s
</usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmbx12.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmex10.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi12.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmmi8.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr12.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmr8.pfb></usr/share/texmf-dist/fonts/type1/public/amsfonts/cm/cmsy10.pfb></usr/share/texmf-dist/fonts/type1/urw/times/utmb8a.pfb></usr/share/texmf-dist/fonts/type1/urw/times/utmr8a.pfb></usr/share/texmf-dist/fonts/type1/urw/times/utmri8a.pfb>
Output written on EEMLA.pdf (5 pages, 109428 bytes).
PDF statistics:
52 PDF objects out of 1000 (max. 8388607)
31 compressed objects within 1 object stream
71 PDF objects out of 1000 (max. 8388607)
43 compressed objects within 1 object stream
0 named destinations out of 1000 (max. 500000)
1 words of extra memory for PDF output out of 10000 (max. 10000000)

BIN
EEMLA.pdf
View File

Binary file not shown.

BIN
EEMLA.synctex.gz
View File

Binary file not shown.

21
EEMLA.tex
View File

@@ -80,6 +80,27 @@ Artificial Intelligence (AI) has surged in popularity and capability in recent y
\vec{y} = f\Bigl( \Bigl[ \sum_{j=1}^{n} w_{1j}x_j + b_1,\;\sum_{j=1}^{n} w_{2j}x_j + b_2,\;\ldots,\;\sum_{j=1}^{n} w_{mj}x_j + b_m \Bigr]^T \Bigr)
\]
The vector $\vec{y}$ is the output vector for any given preceding layer, resulting in an $m$ dimensional evaluation. Each connection between layers of an MLP has two values associated: a weight and a bias. Like the resistance across a synapse in the human brain, biases modulate the activation of each neuron. Mathematically, the biases for each connection are valued for each proceeding neuron, resulting in a $m\times 1$ shape. Weights are defined for each neuron, allowing the creation of a matrix $\mathbf{W}$, which is then computed against the activated values of the preceding layer in a matrix multiplication, then summed with the corresponding bias to generate a scalar for each neuron of the next layer. This process is called forward propagation, as compared to backpropagation. This entire propagation is composited inside a vector-valued activation function $\vec{f}$, which is essential in setting the bounds of the system. However, this can be problematic in complex situations.
The end goal of any machine learning experiment is to function like a regression, except with many more parameters. Backpropogation is the process of updating these weights and biases using discrete evaluation to minimize a cost function $L$. Truly, MLPs are the largest optimization problem ever created. In other words, MLPs can be trained to create certain output for given input using the forward propagation to evaluation, a loss function to find error, and a backpropagation optimizer to update its variables.
In recent years, a specialized kind of Machine Learning models have hit the market --- Large Language Models (LLMs). An LLM is a generative MLP that creates human-readable text given a prompt. They stem from early attempts in the 2000s to use neural networks with Recurrent Neural Networks to analyze sequences of words for sentiment, keywords, and grammar \parencite[2]{Wang2024}. With the rise of tokenizers, or models that convert words to vector embeddings that help encode the meaning thereof, the 2017 paper \textit{Attention is All You Need} changed the landscape of AI forever with the introduction of the self-attention transformer. This development allowed models to understand the relationships between words with far less training. While any task is possible --- theoretically --- with neural networks, optimizations such as these allow for lower error with far less training, making the while process more sustainable. The probability of reaching a minimum of the loss function is far greater with such improvements to MLP architecture \parencite[2]{vaswani2023attentionneed}.
These techniques were later commercialized with the advent of GPT-2, GPT-3, and BERT from AI labs like OpenAI and Google's DeepMind \parencite[3]{Wang2024}. With increased supply of Graphical Processing Units (GPUs) and Tensor Processing Units (TPUs), these models began snowballing in scale. This was especially evident starting in 2019 with an iteration of GPT-2 being released with a production size of 1.5 billion parameters. In 2020, GPT-3 scaled up to 175 billion parameters --- achieving true coherence in reasoning for the first time ever for a machine. GPT-4 was released by OpenAI in 2023, with an undisclosed scale in the trillions of parameters. Development investment also climbed into the hundreds of billions of dollars, with new firms such as Anthropic, Grok, etc. Open sourced projects also gained popularity, some backed by multi-billion dollar R\&D teams such as Meta's Llama series.
Functionally, there is no fundamental algorithmic difference between generative and classification models. Indeed, most LLMs are initially trained to generate new sequences of words by setting the loss function to expect the next word in the series of an existing corpus, through a process known as Casual Language Modeling (CLM). For the purposes of commercialization, they have been re-purposed to be prompted as chat-bots by users. This is done by performing backpropagation based on the generation of conversational sequences, with the LLM often instructed to act as if filling out a conversation's transcript.
Several underlying technologies are involved in the lifecycle of an LLM. The process of creating one usually starts with the definition of a vocabulary. Sequences of language are broken into tokens by algorithms called tokenizers. Tokenizers split text into smaller units, which are then encoded into a vector by another MLP. This is done to develop a sense of meaning via the mathematical similarity of similar words. The similarity of two vectors can be calculated using the cosine-similarity formula, which calculates the angle $\phi$ between two vectors.
\[
\cos\phi=\frac{\vec{A}\cdot\vec{B}}{||\vec{A}||||\vec{B}||}
\]
Efforts to increase the performance of LLMs tend to include provisions for an increased vocabulary of cardinal tokens, leading to more efficient generation of text since more complex words, numbers, and symbols would normally need multiple tokens with the use of techniques like Byte Pair Encoding.
%define LLM benchmarking
Benchmarks for evaluating Large Language Models (LLMs) assess their performance across various tasks, including reasoning, comprehension, generation, and factual accuracy. Standard benchmarks include GLUE and SuperGLUE for natural language understanding, MMLU (Massive Multitask Language Understanding) for evaluating knowledge across diverse subjects, and BIG-bench for measuring reasoning and generalization capabilities \parencite[8]{ivanov2024}. HELLASWAG and LAMBADA test commonsense reasoning and long-range dependency understanding, while TruthfulQA and BBQ assess biases, factual consistency, and ethical alignment \parencite[6]{ivanov2024}. Additionally, human evaluations and BLEU, ROUGE, and METEOR scores help measure text generation quality. As LLMs advance, new benchmarks continuously emerge to capture nuances in performance, efficiency, and ethical behavior.
Adding to the complexity of creating increasingly more performant are the computational and capital costs of building AI-capable supercomputers, clusters, and data centers for corpora, or CLM text databases. Improvements in model architecture are sought before attempts to increase the scale of models and their parameter counts because of the prohibitive scaling laws of neural networks.
%%%%Works cited

22
references.bib
View File

@@ -12,7 +12,25 @@ ISSN = {2076-3417},
ABSTRACT = {Creating learning models that can exhibit sophisticated reasoning abilities is one of the greatest challenges in deep learning research, and mathematics is rapidly becoming one of the target domains for assessing scientific progress in this direction. In the past few years there has been an explosion of neural network architectures, datasets, and benchmarks specifically designed to tackle mathematical problems, reporting impressive achievements in disparate fields such as automated theorem proving, numerical integration, and the discovery of new conjectures or matrix multiplication algorithms. However, despite this notable success it is still unclear whether deep learning models possess an elementary understanding of quantities and numbers. This survey critically examines the recent literature, concluding that even state-of-the-art architectures and large language models often fall short when probed with relatively simple tasks designed to test basic numerical and arithmetic knowledge.},
DOI = {10.3390/app14020744}
}
@article{Wang2024, title={History, development, and principles of large language models: An introductory survey}, DOI={10.1007/s43681-024-00583-7}, journal={AI and Ethics}, author={Wang, Zichong and Chu, Zhibo and Doan, Thang Viet and Ni, Shiwen and Yang, Min and Zhang, Wenbin}, year={2024}, month={Oct}}
@misc{vaswani2023attentionneed,
title={Attention Is All You Need},
author={Ashish Vaswani and Noam Shazeer and Niki Parmar and Jakob Uszkoreit and Llion Jones and Aidan N. Gomez and Lukasz Kaiser and Illia Polosukhin},
year={2023},
eprint={1706.03762},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/1706.03762},
}
@misc{ivanov2024,
title={AI Benchmarks and Datasets for LLM Evaluation},
author={Todor Ivanov and Valeri Penchev},
year={2024},
eprint={2412.01020},
archivePrefix={arXiv},
primaryClass={cs.DC},
url={https://arxiv.org/abs/2412.01020},
}