An Approach to Solving the Abstraction and Reasoning Corpus (ARC) Challenge
This work addresses the ARC challenge, a benchmark for AI reasoning, but is incremental as it builds on existing LLM capabilities with minor tweaks.
The authors tackled the Abstraction and Reasoning Corpus (ARC) challenge by using prompt engineering with GPT-4 to generate descriptions and steps for solving tasks, achieving 2 out of 4 successes on small grid challenges and suggesting improvements could increase this.
We utilise the power of Large Language Models (LLMs), in particular GPT4, to be prompt engineered into performing an arbitrary task. Here, we give the model some human priors via text, along with some typical procedures for solving the ARC tasks, and ask it to generate the i) broad description of the input-output relation, ii) detailed steps of the input-output mapping, iii) use the detailed steps to perform manipulation on the test input and derive the test output. The current GPT3.5/GPT4 prompt solves 2 out of 4 tested small ARC challenges (those with small grids of 8x8 and below). With tweaks to the prompt to make it more specific for the use case, it can solve more. We posit that when scaled to a multi-agent system with usage of past memory and equipped with an image interpretation tool via Visual Question Answering, we may actually be able to solve the majority of the ARC challenge