System 2 Reasoning for Human-AI Alignment: Generality and Adaptivity via ARC-AGI

TL;DR

This paper identifies limitations of transformer models in System 2 reasoning for human-AI alignment and proposes a new framework with symbolic representation, feedback loops, and task augmentation to improve generality and adaptivity.

Contribution

It introduces three research axes—symbolic pipeline, interactive feedback, and test-time augmentation—to enhance reasoning capabilities for better human-AI alignment.

Findings

ARC-AGI exposes gaps in compositional generalization and rule adaptation.

Proposed framework improves reasoning robustness and adaptability.

Evaluation suite tracks progress in symbolic generality and task robustness.

Abstract

Despite their broad applicability, transformer-based models still fall short in System~2 reasoning, lacking the generality and adaptivity needed for human--AI alignment. We examine weaknesses on ARC-AGI tasks, revealing gaps in compositional generalization and novel-rule adaptation, and argue that closing these gaps requires overhauling the reasoning pipeline and its evaluation. We propose three research axes: (1) Symbolic representation pipeline for compositional generality, (2) Interactive feedback-driven reasoning loop for adaptivity, and (3) Test-time task augmentation balancing both qualities. Finally, we demonstrate how ARC-AGI's evaluation suite can be adapted to track progress in symbolic generality, feedback-driven adaptivity, and task-level robustness, thereby guiding future work on robust human--AI alignment.