On Discovering Interesting Combinatorial Integer Sequences

Authors: Martin Svatoš, Peter Jung, Jan Tóth, Yuyi Wang, Ondřej Kuželka

IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We ran two separate experiments with generators of FO2 and C2 sentences. We set a time-limit of five minutes for the computation of combinatorial spectra per sentence; results for these experiments are depicted in Figure 1. Our aim with these experiments was to assess the effect of the pruning techniques that we proposed.
Researcher Affiliation Collaboration Martin Svatoˇs1 , Peter Jung1 , Jan T oth1 , Yuyi Wang2,3 and Ondˇrej Kuˇzelka1 1Faculty of Electrical Engineering, Czech Technical University in Prague, Czech Republic 2CRRC Zhuzhou Institute, China 3ETH Zurich, Switzerland
Pseudocode No The paper describes algorithms in prose and provides examples, but it does not include any formally presented pseudocode blocks or figures labeled 'Algorithm'.
Open Source Code Yes This paper is accompanied by a technical report available at https://arxiv.org/abs/2302.04606 and codes available at https://github.com/martinsvat/Sentence Finder.
Open Datasets Yes For each of the sequences in our database, we queried OEIS to determine if the sequence matches a sequence which is already in OEIS. We found that 301 of the sequences were present in OEIS this makes 1.2% of the sequences we generated. ...OEIS Foundation Inc. The on-line encyclopedia of integer sequences. Published electronically at http://oeis.org, 2023. Accessed: 2023-05-23.
Dataset Splits No The paper describes its experimental process, but it does not explicitly define or mention training, validation, or test dataset splits in the context of preparing data for a model.
Hardware Specification Yes For that we let the sentence generator run for five days to obtain a collection of sentences and their combinatorial spectra on a machine with 500 GB RAM, 128 processors (we used multi-threading).
Software Dependencies No The paper mentions using 'our implementation of the algorithm from [van Bremen and Kuˇzelka, 2021]' and 'our implementation of the reductions from [Kuˇzelka, 2021]', but it does not provide specific version numbers for any software libraries, frameworks, or programming languages.
Experiment Setup Yes We set a time-limit of five minutes for the computation of combinatorial spectra per sentence... For that we let the sentence generator run for five days...