# Multi-Perspective Transformers in ARC-AGI-2 Challenge

arXiv: https://arxiv.org/abs/2605.01154 · HTML: https://mlsanity.com/paper/2605.01154/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 2 of its 5 references that are in the map.
- First posted on arXiv 1 May 2026.

## Builds on (most cited first, 2 of 2)

- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Product of Experts with LLMs: Boosting Performance on ARC Is a Matter of Perspective](https://mlsanity.com/paper/2505.07859/) (2025)

## Abstract (arXiv)

ARC-AGI-2 is a benchmark of human-intuitive visual puzzles that measures a machine's ability to generalize from limited examples, interpret symbolic meaning, and flexibly apply rules in varying contexts. In this paper, we discuss our approach to solving the ARC-AGI-2 puzzles with TinyLM, with additional fine-tuning at test time, including Test-Time-Training (TTT) and Products of Experts (POE). Our model achieves 96.1% accuracy on the training set and 21.7% accuracy on the evaluation set.