Evolving Minds: Logic-Informed Inference from Temporal Action Patterns

Authors: Chao Yang, Shuting Cui, Yang Yang, Shuang Li

ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on both synthetic and real-world datasets show that our method outperforms existing approaches in accurately inferring mental states and predicting actions, demonstrating its effectiveness in modeling human cognitive processes. (Abstract) ... 5. Experiments ... 5.1. Experimental Setup ... 5.2. Experiments on Synthetic Dataset ... 5.3. Experiments on Real-World Dataset
Researcher Affiliation Academia 1School of Data Science, The Chinese University of Hong Kong (Shenzhen), Shenzhen, China. Correspondence to: Shuang Li <EMAIL>.
Pseudocode Yes Algorithm 1 Predicting Actions with Backtracking ... Alg. 2 in Appendix. A.1, using inverse transform sampling method. ... The algorithm for the column generation is presented in Alg. 3, Appendix. A.2. ... algorithm shown in Alg. 4, Appendix. A.3. ... In Alg. 5, we provide the pseudocodes for the backtracking sampling mechanism
Open Source Code No The paper does not contain any explicit statements or links indicating that source code for the described methodology is publicly available.
Open Datasets Yes Hand Me-That (Wan et al., 2022)... Car-following (Li et al., 2023)... Multi THUMOS (Yeung et al., 2018)... EPIC-Kitchen-100 (Damen et al., 2018)
Dataset Splits Yes For all the datasets, we split each dataset into 80%, 10%, 10% train/dev/test by the total sequences.
Hardware Specification Yes All synthetic data experiments and real-world data experiments, including the comparison experiments with baselines, are performed on Ubuntu 20.04.3 LTS system with Intel(R) Xeon(R) Gold 6248R CPU @ 3.00GHz, 227 Gigabyte memory.
Software Dependencies No The paper mentions the operating system (Ubuntu 20.04.3 LTS) but does not provide specific version numbers for any programming languages, libraries, or other software dependencies critical for replication.
Experiment Setup Yes We present the selected hyper-parameters on both synthetic datasets and real-world datasets in Tab. 19. (Section E.1: Hyper-Parameter Selection) ... Table 19. Descriptions and values of hyper-parameters used for models trained on both synthetic dataset and real-world datasets.