Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Reactive Dialectic Search Portfolios for MaxSAT
Authors: Carlos Ansยtegui, Josep Pon, Meinolf Sellmann, Kevin Tierney
AAAI 2017 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We apply it to Max SAT and compare it empirically with state-of-the-art Max SAT solvers as well as its non-reactive counterpart, both in combination with the existing solver and in isolation. In the end, we obtain a solver that outperforms the state of the art in various categories of heuristic Max SAT solving, as assessed independently in the 2016 Max SAT Evaluation (Argelich et al. 2016). Numerical Results Having developed our approach in the previous section, we now evaluate it empirically. |
| Researcher Affiliation | Collaboration | Carlos Ans otegui, Josep Pon DIEI Universitat de Lleida, Spain EMAIL Meinolf Sellmann IBM Research, USA EMAIL Kevin Tierney DS&OR Lab University of Paderborn, Germany EMAIL |
| Pseudocode | Yes | Algorithm 1 Parameterized Dialectic Search |
| Open Source Code | No | The paper does not provide any statement about releasing source code or a link to a code repository for their described methodology. |
| Open Datasets | Yes | Our base set of Max SAT instances are all instances in the Random and Crafted categories in the Max SAT Evaluation 2016 (MSE16) (Argelich et al. 2016). Argelich, J.; Li, C.; Many a, F.; and Planes, J. 2016. Max SAT Evaluations. www.maxsat.udl.cat. |
| Dataset Splits | No | The paper states, 'We cleanly split each group randomly 80 to 20, whereby the 80% are assigned to our training set while the remaining 20% are set aside for testing.' It explicitly mentions training and testing sets but no separate validation set. |
| Hardware Specification | Yes | We run all our experiments on a cluster featured with Intel Xeon CPU E5-26020 @ 2.6GHz processors, a memory limit of 3.5 GB, and each machine runs an instance of Rocks Cluster 6.5 (Linux 2.6.32), which is the exact same environment used in the MSE16. |
| Software Dependencies | Yes | each machine runs an instance of Rocks Cluster 6.5 (Linux 2.6.32) |
| Experiment Setup | Yes | We use a distributed version of GGA++ with 8 machines with 8 cores each, a population size of 100 individuals and 100 generations, using a 30 second target algorithm timeout. The time limit for the test instances is as in the MSE16, 300 seconds. |