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].
Sequential Monte Carlo for Graphical Models
Authors: Christian Andersson Naesseth, Fredrik Lindsten, Thomas B Schön
NeurIPS 2014 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section we evaluate the proposed SMC sampler on three examples to illustrate the merits of our approach. Additional details and results are available in the supplementary material and code to reproduce results can be found in [27]. We first consider an example from statistical mechanics, the classical XY model, to illustrate the impact of the sequential decomposition. Furthermore, we profile our algorithm with the gold standard AIS [2] and Annealed Sequential Importance Resampling (ASIR1) [11]. |
| Researcher Affiliation | Academia | Christian A. Naesseth Div. of Automatic Control Link oping University Link oping, Sweden EMAIL Fredrik Lindsten Dept. of Engineering The University of Cambridge Cambridge, UK EMAIL Thomas B. Sch on Dept. of Information Technology Uppsala University Uppsala, Sweden EMAIL |
| Pseudocode | Yes | Algorithm 1 Sequential Monte Carlo (SMC) |
| Open Source Code | Yes | Additional details and results are available in the supplementary material and code to reproduce results can be found in [27]. Reference [27]: C. A. Naesseth, F. Lindsten, and T. B. Sch on. smc-pgm, 2014. URL http://dx.doi.org/10. 5281/zenodo.11947. |
| Open Datasets | Yes | We use the same data and learnt models as Wallach et al. [15], i.e. 20 newsgroups, and Pub Med Central abstracts (PMC). |
| Dataset Splits | No | The paper mentions 'held-out documents' for evaluation but does not provide specific percentages or sample counts for training, validation, and test splits across all experiments, nor does it refer to predefined standard splits with citations that would enable reproduction of data partitioning. |
| Hardware Specification | No | The paper does not provide specific hardware details such as exact GPU/CPU models, processor types, or memory amounts used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific software details like library or solver names with version numbers needed to replicate the experiment. |
| Experiment Setup | Yes | We show results for the following cases: Random neighbour (RND-N) First node selected randomly among all nodes, concurrent nodes selected randomly from the set of nodes with a neighbour in XLk 1. Diagonal (DIAG) Nodes added by traversing diagonally (45 angle) from left to right. Spiral (SPIRAL) Nodes added spiralling in towards the middle from the edges. Left-Right (L-R) Nodes added by traversing the graph left to right, from top to bottom. We also give results of AIS with single-site-Gibbs updates and 1 000 annealing distributions linearly spaced from zero to one, starting from a uniform distribution (geometric spacing did not yield any improvement over linear spacing for this case). We set N = 105 for the proposed SMC sampler and then match the computational costs of AIS and ASIR with this computational budget. We use a moderate number of N = 50 particles in the PMCMC sampler. |