ARCLE: The Abstract and Reasoning Corpus Learning Environment for Reinforcement Learning
Published in CoLLAs 2024, 2024
This paper is about ARCLE, environment for reinforcement learning to solve ARC task
Publications
Reasoning Abilities of Large Language Models: In-Depth Analysis on the Abstraction and Reasoning Corpus
Published in ACM TIST submitted, 2024
The existing methods for evaluating the inference abilities of Large Language Models (LLMs) have been results-centric, making it difficult to assess the inference process. We introduce a new approach using the Abstract and Reasoning Corpus (ARC) dataset to evaluate the inference and contextual understanding abilities of large language models in a process-centric manner. ARC demands rigorous logical structures for problem-solving, making it a benchmark that facilitates the comparison of model inference abilities with humans. Experimental results confirm that while large language models possess weak inference abilities, they still lag in terms of logical coherence, compositionality, and productivity. Our experiments highlight the reasoning capabilities of LLMs, proposing development paths for achieving human-level reasoning.
Recommended citation: https://arxiv.org/abs/2403.11793
Extracting the core knowledge of ARC with the World Model
Published in KSC, 2023
In this study, we design an experiment to verify that the World Model extracts core knowledge from the ARC dataset, and propose future research directions for utilizing this extracted core knowledge.
Evaluating Prior Knowledge of ARC Using World Models
Published in KSC, 2023
In this study, we propose a new method that uses the World Model algorithm to analyze the influence of prior knowledge when solving the ARC benchmark, as well as the types of prior knowledge included in the ARC.