%5Ko7J=z~u>Aj9@bvKf:B%t cz7t b+]h6g%b 6iDj01l4mm>2`Dmz֨X.o;m: ,;`2.| d.|q-l8|3f׮0o@o6o7Ff:nCÐ^;tǀ4>Łbl!1!2ζ#$Õ4Š67Ɛ!"Ϫ.ϒ9ɐ-3đbz!$ˈ3tb%Ő25ztcvي5$ɀ0&ޗz?đ%$Ą,t79؉$ˈ%t ڐ0&ϑ%3غ5"É)"Ӻ79؉$ =z$trz%%ٌ/8blqgdwvbscyun=@V7584nP-~/18;##$)%?>r`1>#n r*=)rv3c@dB{7 4bJi{DbxC~P|n" <1!Pjn- 958"9 )" -nP7>$&)< 5?Pj851P9(HaP#)9#;4nP}JfCb{C`zCa1uJyB1r1LrP2{l=^ymk 4L$Czh=_yggEs} S``*S[l=Dj:y`l;Dm,:+9Wdh$E432{f+_zlk ph%Es%kAm=^43x$9e~l Q~}k '9q:+-S``*SEj(ZsO(Ubf;,8eej;SsgDl'Bw} Yx+sM4h'Qzlk &%kBoy,,+9Yd};W}Dd(Do+4K:+ R43x:+:Sez Yx@-,+'1u$8{T8y"8x#J R1 .9uWJkKI6 'm ?(p lvj@dkg'Od vW'g' i)'\)vk ?+4[7_G]1\0/A)=,DOm+$r5"555&9|"r%$ĄʝpȒy 5$˕g͐pp53p?aψ`Έ`6͑~Ȇ{ ĉ~"5?|rϐ|#>jab̑ΐdϕe캓hP2026-01-26 08:09:09 info: [Puppeteer Page] Got cookies, applying... fmZ?hvZxLzZ?]dΆ?KxYzZpZ|X'hOiZ=oxbԺrKԞrPxL|OdV3`B?zw,4a||C}o,"aj4uw~4,L[Ó",}}k,U^~kk|oi.M}zmz.meg.h.~k.S,2026-01-26 08:09:09 info: [Puppeteer Page] Set 0 cookies on the page j8q##V*(dizVM!3dP=rQ2?dzcM7!#(zu[#=#qD`#q5dQ 9fGqb#y-qR6,dG!xQC4=\=ucxbM<3hG xnLs,iGs(`E6z|_TX"Sj8{zǡPqmgmhI{<$gmǡl2<ߪ2mp$ݚ/ݭ,ԩ\ԫ*Ԫ+ْZߞ&_u`Ӂ囫|w%njlb㐣`%~ٹ%䚲bCա@jf=hsiuÐbqD₵k†r'䚲btゥb\3We5PLkE#K5gP.M L"gP<gF }(L$Zd0O1Ze^ 1Z5V"[7\eZ&#W1LjV,$[7V1Z0]'V Z5MjV1G'],Z7Y7Q s8qATE?2026-01-26 08:09:10 info: [Puppeteer File-Downloader] Attempting to download asset directly... 6j89ͳo"ֳ?إ>ڳo"w쟢4waг>ܱ(Ȧ=ɳ(坐$ ʸ!ܲ( ɳ ɿ#"ٹ:ѹ,>آmԤ(ɺ4o6j8Mhttp/1.1}@@proceedings{1164, author = {Kevin Cao and Kevin Leach}, title = {Revisiting Deep Learning for Variable Type Recovery}, abstract = {

Compiled binary executables are often the only available artifact in reverse engineering, malware analysis, or maintenance of software systems. Unfortunately, the lack of semantic information like variable names makes comprehending binaries difficult. In efforts to improve the comprehensibility of binaries, researchers have recently used machine learning techniques to predict semantic information contained in the original source code. Chen et al. implemented DIRTY, a Transformer-based Encoder-Decoder architecture capable of augmenting decompiled code with variable names and types by leveraging decompiler output tokens and variable size information. Chen et al. were able to demonstrate a substantial increase in name and type extraction accuracy on Hex-Rays decomiler outputs compared to existing static analysis and AI-based techniques. We extend the original DIRTY results by re-training the DIRTY model on a dataset produced by the open-source Ghidra decompiler. Although Chen et al. concluded that Ghidra was not a suitable decompiler candidate due to its difficulty in parsing DWARF, we demonstrate that straightforward parsing of variable data generated by Ghidra results in similar retyping performance. We hope this work inspires further interest and adoption of the Ghidra decompiler for use in research projects.

}, year = {2023}, journal = {International Conference on Program Comprehension}, month = {05/2023}, publisher = {ICPC 2023}, address = {Melbourne, Australia}, url = {https://arxiv.org/abs/2304.03854}, doi = {https://doi.org/10.48550/arXiv.2304.03854}, } "t9$%t9l/lz%3п 01v$v5,ʰvĭnثnd3Vՠ3|#J/DsJ5/"\r^uRrUW'TT'dmfFb׆8m9;l4t>xl)pwiso_-i8|o_4<ǭJ3}ssnXk<ƪn(ܷ3;8a(m3Z)|}k5׊/~8w#v8z9;p*n l*h k*s/dِo(|)ז5י8m3[