Don't throw the baby out with the bathwater: How and why deep learning for ARC

• URL: http://arxiv.org/abs/2506.14276v1
• Date: Tue, 17 Jun 2025 07:40:39 GMT
• Title: Don't throw the baby out with the bathwater: How and why deep learning for ARC
• Authors: Jack Cole, Mohamed Osman,
• Abstract summary: Abstraction and Reasoning (ARC-AGI) presents a formidable challenge for AI systems.<n>We propose a methodology for training on ARC, starting from pretrained LLMs, and enhancing their ARC reasoning.<n>We are the first to propose and show deep learning can be used effectively for ARC, showing boosts of up to 260% in accuracy with AIRV and a further 300% boost with TTFT.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The Abstraction and Reasoning Corpus (ARC-AGI) presents a formidable challenge for AI systems. Despite the typically low performance on ARC, the deep learning paradigm remains the most effective known strategy for generating skillful (state-of-the-art) neural networks (NN) across varied modalities and tasks in vision, language etc. The deep learning paradigm has proven to be able to train these skillful neural networks and learn the abstractions needed in these diverse domains. Our work doubles down on that and continues to leverage this paradigm by incorporating on-the-fly NN training at test time. We demonstrate that fully committing to deep learning's capacity to acquire novel abstractions yields state-of-the-art performance on ARC. Specifically, we treat both the neural network and the optimizer (rather than just a pre-trained network) as integral components of the inference process, fostering generalization to unseen tasks. Concretely, we propose a methodology for training on ARC, starting from pretrained LLMs, and enhancing their ARC reasoning. We also propose Test-Time Fine-Tuning (TTFT) and the Augment Inference Reverse-Augmentation and Vote (AIRV) as effective test-time techniques. We are the first to propose and show deep learning can be used effectively for ARC, showing boosts of up to 260% in accuracy with AIRV and a further 300% boost with TTFT. An early version of this approach secured first place in the 2023 ARCathon competition, while the final version achieved the current best score on the ARC private test-set (58%). Our findings highlight the key ingredients of a robust reasoning system in unfamiliar domains, underscoring the central mechanisms that improve broad perceptual reasoning.

Related papers

• Efficient Neural Clause-Selection Reinforcement [0.0]
  We present a neural network architecture for scoring clauses for clause selection.<n>We integrate the network into the Vampire theorem prover and train it from successful proof attempts.<n>An experiment on the diverse TPTP benchmark finds the neurally guided prover improve over a baseline strategy.
  arXiv  Detail & Related papers  (2025-03-10T19:14:48Z)
• Towards Theoretically Inspired Neural Initialization Optimization [66.04735385415427]
  We propose a differentiable quantity, named GradCosine, with theoretical insights to evaluate the initial state of a neural network. We show that both the training and test performance of a network can be improved by maximizing GradCosine under norm constraint. Generalized from the sample-wise analysis into the real batch setting, NIO is able to automatically look for a better initialization with negligible cost.
  arXiv  Detail & Related papers  (2022-10-12T06:49:16Z)
• Neural Architecture Search for Speech Emotion Recognition [72.1966266171951]
  We propose to apply neural architecture search (NAS) techniques to automatically configure the SER models. We show that NAS can improve SER performance (54.89% to 56.28%) while maintaining model parameter sizes.
  arXiv  Detail & Related papers  (2022-03-31T10:16:10Z)
• Efficient Few-Shot Object Detection via Knowledge Inheritance [62.36414544915032]
  Few-shot object detection (FSOD) aims at learning a generic detector that can adapt to unseen tasks with scarce training samples. We present an efficient pretrain-transfer framework (PTF) baseline with no computational increment. We also propose an adaptive length re-scaling (ALR) strategy to alleviate the vector length inconsistency between the predicted novel weights and the pretrained base weights.
  arXiv  Detail & Related papers  (2022-03-23T06:24:31Z)
• Human-Level Control through Directly-Trained Deep Spiking Q-Networks [16.268397551693862]
  Spiking Neural Networks (SNNs) have great potential on neuromorphic hardware because of their high energy-efficiency. We propose a directly-trained deep spiking reinforcement learning architecture based on the Leaky Integrate-and-Fire neurons and Deep Q-Network. Our work is the first one to achieve state-of-the-art performance on multiple Atari games with the directly-trained SNN.
  arXiv  Detail & Related papers  (2021-12-13T09:46:17Z)
• Improving the sample-efficiency of neural architecture search with reinforcement learning [0.0]
  In this work, we would like to contribute to the area of Automated Machine Learning (AutoML) Our focus is on one of the most promising research directions, reinforcement learning. The validation accuracies of the child networks serve as a reward signal for training the controller. We propose to modify this to a more modern and complex algorithm, PPO, which has demonstrated to be faster and more stable in other environments.
  arXiv  Detail & Related papers  (2021-10-13T14:30:09Z)
• A Simple Fine-tuning Is All You Need: Towards Robust Deep Learning Via Adversarial Fine-tuning [90.44219200633286]
  We propose a simple yet very effective adversarial fine-tuning approach based on a $textitslow start, fast decay$ learning rate scheduling strategy. Experimental results show that the proposed adversarial fine-tuning approach outperforms the state-of-the-art methods on CIFAR-10, CIFAR-100 and ImageNet datasets.
  arXiv  Detail & Related papers  (2020-12-25T20:50:15Z)
• Continual Learning for Natural Language Generation in Task-oriented Dialog Systems [72.92029584113676]
  Natural language generation (NLG) is an essential component of task-oriented dialog systems. We study NLG in a "continual learning" setting to expand its knowledge to new domains or functionalities incrementally. The major challenge towards this goal is catastrophic forgetting, meaning that a continually trained model tends to forget the knowledge it has learned before.
  arXiv  Detail & Related papers  (2020-10-02T10:32:29Z)
• Unbiased Deep Reinforcement Learning: A General Training Framework for Existing and Future Algorithms [3.7050607140679026]
  We propose a novel training framework that is conceptually comprehensible and potentially easy to be generalized to all feasible algorithms for reinforcement learning. We employ Monte-carlo sampling to achieve raw data inputs, and train them in batch to achieve Markov decision process sequences. We propose several algorithms embedded with our new framework to deal with typical discrete and continuous scenarios.
  arXiv  Detail & Related papers  (2020-05-12T01:51:08Z)
• Improving Performance in Reinforcement Learning by Breaking Generalization in Neural Networks [5.273501657421096]
  We show how online NN training and interference interact in reinforcement learning. We find that simply re-mapping the input observations to a high-dimensional space improves learning speed and parameter sensitivity. We provide a simple approach to NN training that is easy to implement, and requires little additional computation.
  arXiv  Detail & Related papers  (2020-03-16T19:21:08Z)
• Robust Pruning at Initialization [61.30574156442608]
  A growing need for smaller, energy-efficient, neural networks to be able to use machine learning applications on devices with limited computational resources. For Deep NNs, such procedures remain unsatisfactory as the resulting pruned networks can be difficult to train and, for instance, they do not prevent one layer from being fully pruned.
  arXiv  Detail & Related papers  (2020-02-19T17:09:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.