Bayesian Affect Control Theory of Self
Authors: Jesse Hoey, Tobias Schroeder
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Simulations To implement Bayes Act-S in practice, we make use of the Bayes Act function SIMULATE(fa, fc) (f a, f c)5. This function is used to map a distribution over agent and client identities, fa and fc, respectively, to another such distribution after a single interaction (agent acts, client acts). ... Figure 1 shows the results of a simulation with T = 20, η = 0.95, ηf = 1.0 and with N = 2000 samples for the identities in Bayes Act. |
| Researcher Affiliation | Academia | Jesse Hoey David R. Cheriton School of Computer Science University of Waterloo, 200 University Ave. West, Waterloo, Ontario, N2L3G1, Canada jhoey@cs.uwaterloo.ca Tobias Schr oder Potsdam University of Applied Sciences, Kiepenheuerallee 5, 14469 Potsdam, Germany post@tobiasschroeder.de |
| Pseudocode | No | The paper describes methods and models using natural language and mathematical equations but does not include any formal pseudocode or algorithm blocks. |
| Open Source Code | Yes | Bayes Act code and videos of simulations are available at bayesact.ca. |
| Open Datasets | Yes | Affect control theorists have compiled lexicons of a few thousand words along with average EPA ratings obtained from survey participants who are knowledgeable about their culture (Heise 2010). For example, most English speakers agree that professors are about as nice as students (E), more powerful (P) and less active (A). The corresponding EPAs are [1.7, 1.8, 0.5] for professor and [1.8, 0.7, 1.2] for student2. |
| Dataset Splits | No | The paper describes simulation parameters and initial conditions but does not specify training, validation, or test dataset splits, nor does it describe cross-validation or other data partitioning methods. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware (CPU, GPU, memory, or specific computer models) used for running the simulations or experiments. |
| Software Dependencies | No | The paper mentions using 'Bayes Act code' and a 'SIMULATE' function, but it does not specify any software dependencies with version numbers (e.g., programming language versions, library versions). |
| Experiment Setup | Yes | Figure 1 shows the results of a simulation with T = 20, η = 0.95, ηf = 1.0 and with N = 2000 samples for the identities in Bayes Act. The Bayes Act parameters are: Σ is diagonal with elements 0.1, and Σf is diagonal with elements 0.001 for client identity, and 0.01 for agent identity. ... The initial variances on agent and client identities are 0.01. |