Path: utzoo!attcan!uunet!tut.cis.ohio-state.edu!ucbvax!pasteur!agate!shelby!portia!portia.stanford.edu!bugboy From: bugboy@portia.Stanford.EDU (Michael Frank) Newsgroups: comp.ai.neural-nets Subject: Re: a NN for learning UNIX programs Message-ID: <10398@portia.Stanford.EDU> Date: 21 Mar 90 16:52:09 GMT References: <12724@csli.Stanford.EDU> Sender: USENET News System Reply-To: bugboy@portia.Stanford.EDU (Michael Frank) Organization: Stanford University Lines: 33 In <12724@csli.Stanford.EDU> I wrote: >The following is the abstract of my final project for David >Rumelhart's introductory class here at Stanford. Anyone interested, >send me email, and I'll send you the rest of the paper, and source for >the program (in UNIX C with Curses) if you so desire. > A user-friendly program was created to connect a recurrent PDP >network to arbitrary UNIX programs, in such a way that the network >could learn through back-propagation to predict the program's textual >response to characters in an input stream. In numerous experiments, >the network was able to learn, for an impressive variety of simple >programs and input regimens, to predict exactly what the next text >character produced by the UNIX program would be. Results of these >experiments are described, and a proposal is made for future work that >would see if a PDP network could be made to explore complex UNIX >programs on its own. The UNIX environment is promoted as a good >testbed "world" that proposed mind-emulating programs could learn to >explore. Due to the large number of requests I've received for the paper and program, I'm making them available for anonymous FTP from csli.Stanford.EDU. They are in /usr/ftp/pub/neunix. E.g., ftp csli.stanford.edu anonymous anonymous cd pub/neunix mget * Thank you all for your interest! , , __ /|/| . _ |_ _ _ | |_ _ _ ,_ |, / | | | (_ | | (_| (-' | | | (_| | | |\ mpf@csli.stanford.edu bugboy@portia.stanford.edu bugboy%portia@stanford.bitnet