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 embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..

Analyzing the Power of Chain of Thought through Memorization Capabilities

Authors: Lijia Yu, Xiao-Shan Gao, Lijun Zhang

NeurIPS 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Our contributions are theoretical research.
Researcher Affiliation Academia 1Institute of AI for Industries, Chinese Academy of Sciences 2SKLMS, Academy of Mathematics and Systems Science, Chinese Academy of Sciences 3University of Chinese Academy of Sciences 4Key Laboratory of System Software of Chinese Academy of Sciences
Pseudocode No The paper includes mathematical definitions, theorems, lemmas, and proofs, but no clearly labeled pseudocode or algorithm blocks describing a computational procedure.
Open Source Code No Our contributions are theoretical research.
Open Datasets No Our contributions are theoretical research.
Dataset Splits No The paper does not describe any experimental datasets or their splits, as it is theoretical research. It discusses conceptual 'finite reasoning datasets' and 'infinite reasoning datasets'.
Hardware Specification No Our contributions are theoretical research.
Software Dependencies No Our contributions are theoretical research.
Experiment Setup No Our contributions are theoretical research.