Generative Language Modeling for Automated Theorem Proving · HackerPapers discussions

Generative Language Modeling for Automated Theorem Proving — mentioned by tsoj

Thu, 25 Jul 2024 16:38:47 GMT

This is NOT the paper, but probably a very similar solution: https://arxiv.org/abs/2009.03393

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by aSanchezStern

Fri, 27 Oct 2023 18:52:58 GMT

Sounds like you've stumbled into the wonderful world of machine-learning guided proof synthesis! While I don't think the full system you're describing has been built yet, many similar systems and pieces have. In terms of the initial phase of supervised learning on existing proofs to prove new ones, there's TacticToe (https://arxiv.org/abs/1804.00596), Tactician (https://arxiv.org/pdf/2008.00120.pdf), CoqGym/ASTactic (https://arxiv.org/abs/1905.09381), Proverbot9001 (https://arxiv.org/abs/1907.07794), and Diva (https://dl.acm.org/doi/10.1145/3510003.3510138#sec-terms), among others. Most of…

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by gwern

Tue, 14 Jun 2022 17:27:53 GMT

Because Nethack and video games remove most of the reasons that GOFAI failed; for example, everything hard about perception from pixels is removed when you have a little parsed text grid of cells presented to you or are hooked into the game engine and query objects directly. (The real world is not made of a clean little grid of discrete high-level objects like 'dragon' or 'black jelly'.) Moravec's paradox again: humans find perception easy but things like Sokoban hard, while GOFAI approaches find Sokoban so trivial that it's common to use it (or Sudoku) as a toy problem introduction to…

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by gwern

Sat, 19 Jun 2021 18:32:47 GMT

It's worth noting that GPT-f already gets a big performance boost from pretraining on Arxiv etc (https://arxiv.org/pdf/2009.03393.pdf#page=7) despite those sources containing next to no Metamath or anything that looks like a raw Metamath proof, just regular natural language & LaTeX discussing math...

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by Veedrac

Thu, 07 Jan 2021 16:13:23 GMT

> The accuracy largely comes from the fact that addition rarely requires carrying more than a single digit. So it's easy to pattern match from single digit problems that it was previously trained on. Again, not at all true due to BPEs. [12] [65] [25] [42] [185] + [580] [22] [75] [80] [93] --------------------------- [706] [75] [300] [278] (note that GPT-3 is never told that the BPE [580] is composed of the digits [5], [8], and [0]. It has to guess this from the contexts [580] occurs in.) > With multiplication, which requires much more extensive cross-column interaction, accuracy falls off a…

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by Veedrac

Wed, 11 Nov 2020 02:58:40 GMT

> Aren't you shifting the goalposts a bit? No, I don't think so. My original comment was “I have looked at the program synthesis literature before and it really does not seem very advanced to me. The General Program Synthesis Benchmark Suite lists unsolved benchmarks like “Given three strings n1, n2, and n3, return true if length(n1) < length(n2) < length(n3), and false otherwise”, and that's with 100 examples. So, probably less practically useful than GPT-3, which wasn't even trained on the task.” I only asked about ILP to clarify your defense. I maintain, after seeing the ILP you were…

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by 7373737373

Wed, 30 Sep 2020 14:21:47 GMT

Maybe ask GPT-f (https://arxiv.org/abs/2009.03393) next time!

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by gwern

Mon, 21 Sep 2020 19:14:14 GMT

OA has done that, as 'GPT-f': https://arxiv.org/abs/2009.03393#openai They also have ambitions of IMO: https://twitter.com/spolu/status/1303693595833237504 I assume it's not mentioned in OP because GPT-f came out too recently for them to interview or profile compared to this Lean-using team.

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — mentioned by thomasahle

Wed, 09 Sep 2020 15:49:04 GMT

Actually OpenAI just published an article about using Transformer for theorem proving: https://arxiv.org/abs/2009.03393

Paper and discussion history. A mention is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — HN submission

Wed, 09 Sep 2020 13:41:50 GMT

Generative Language Modeling for Automated Theorem Proving

88 points; 28 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Generative Language Modeling for Automated Theorem Proving — HN submission

Wed, 09 Sep 2020 01:53:27 GMT

Generative Language Modeling for Automated Theorem Proving

3 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.