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 Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The Shift Toward Open and Reproducible AI Research," Under Review, 2026. doi: 10.48550/arXiv.2606.16974..
Decomposition Bounds for Marginal MAP
Authors: Wei Ping, Qiang Liu, Alexander T. Ihler
NeurIPS 2015 | Venue PDF | LLM Run Details | Input Tokens: 15,846 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 5,993 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We report experimental results in Section 6 and conclude the paper in Section 7. ... In this section, we demonstrate our algorithm on a set of real-world graphical models from recent UAI inference challenges... Figure 2 and Figure 3 compare the convergence of the different algorithms... |
| Researcher Affiliation | Academia | Computer Science, UC Irvine Computer Science, Dartmouth College EMAIL EMAIL |
| Pseudocode | Yes | Algorithm 1 Generalized Dual-decomposition (GDD) |
| Open Source Code | No | No statement regarding the release of the paper's source code is found. |
| Open Datasets | Yes | In this section, we demonstrate our algorithm on a set of real-world graphical models from recent UAI inference challenges, including two diagnostic Bayesian networks with 203 and 359 variables and max domain sizes 7 and 6, respectively, and several MRFs for pedigree analysis with up to 1289 variables, max domain size of 7 and clique size 5.7 We construct marginal MAP problems on these models by randomly selecting half of the variables to be max nodes, and the rest as sum nodes. ... 7See http://graphmod.ics.uci.edu/uai08/Evaluation/Report/Benchmarks. |
| Dataset Splits | No | The paper uses real-world graphical models from UAI inference challenges but does not specify the training/validation/test splits, only that half of the variables are randomly selected as max nodes and the rest as sum nodes. |
| Hardware Specification | No | No specific hardware details (e.g., GPU/CPU models, memory) used for running the experiments are mentioned in the paper. |
| Software Dependencies | No | The paper mentions implementing algorithms and using an 'off-the-shelf L-BFGS implementation', but does not provide specific version numbers for any software dependencies. |
| Experiment Setup | Yes | We implement several algorithms that optimize the same primal marginal MAP bound, including our GDD (Algorithm 1), the WMB algorithm in [16] with ibound = 1... We tune the damping rate of WMB from 0.01 to 0.05... The elimination order that we use is obtained by a weighted-min-๏ฌll heuristic [1] constrained to eliminate the sum nodes ๏ฌrst... In our implementation, we ๏ฌnd that a few gradient steps (e.g., 5) with a backtracking line search using the Armijo rule works well in practice. |