I looked at the one below. This is a bit outside my comfort zone of knowledge. It though looks to be a way to use neural networks to prove theorems.
LC On Saturday, November 7, 2020 at 1:09:48 PM UTC-6 [email protected] wrote: > > Learning Reasoning Strategies in End-to-End Differentiable Proving > https://arxiv.org/abs/2007.06477 > > End-to-End Differentiable Proving > https://arxiv.org/abs/1705.11040 > *End-to-End Differentiable Proving* Tim Rocktäschel <https://arxiv.org/search/cs?searchtype=author&query=Rockt%C3%A4schel%2C+T> , Sebastian Riedel <https://arxiv.org/search/cs?searchtype=author&query=Riedel%2C+S> We introduce neural networks for end-to-end differentiable proving of queries to knowledge bases by operating on dense vector representations of symbols. These neural networks are constructed recursively by taking inspiration from the backward chaining algorithm as used in Prolog. Specifically, we replace symbolic unification with a differentiable computation on vector representations of symbols using a radial basis function kernel, thereby combining symbolic reasoning with learning subsymbolic vector representations. By using gradient descent, the resulting neural network can be trained to infer facts from a given incomplete knowledge base. It learns to (i) place representations of similar symbols in close proximity in a vector space, (ii) make use of such similarities to prove queries, (iii) induce logical rules, and (iv) use provided and induced logical rules for multi-hop reasoning. We demonstrate that this architecture outperforms ComplEx, a state-of-the-art neural link prediction model, on three out of four benchmark knowledge bases while at the same time inducing interpretable function-free first-order logic rules. Comments: NIPS 2017 camera-ready, NIPS 2017 Subjects: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Logic in Computer Science (cs.LO) Cite as: arXiv:1705.11040 <https://arxiv.org/abs/1705.11040> [cs.NE] (or arXiv:1705.11040v2 <https://arxiv.org/abs/1705.11040v2> [cs.NE] for this version) Submission historyFrom: Tim Rocktäschel [view email <https://arxiv.org/show-email/cbb66212/1705.11040>] *[v1] <https://arxiv.org/abs/1705.11040v1>* Wed, 31 May 2017 11:40:57 UTC (32 KB) *[v2]* Mon, 4 Dec 2017 00:24:04 UTC (34 KB) > > Towards Neural Theorem Proving at Scale > https://arxiv.org/abs/1807.08204 > > Neural Theorem Provers Do Not Learn Rules Without Exploration > https://arxiv.org/abs/1906.06805 > > @philipthrift > > -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/82f507b6-9bd4-4bb0-81fd-d68146ecc5ecn%40googlegroups.com.
