paper-with-me

Papers

ARC-AGI Without Pretraining

2025-12-05 · Isaac Liao, Albert Gu arxiv

Conventional wisdom in the age of LLMs dictates that solving IQ-test-like visual puzzles from the ARC-AGI-1 benchmark requires capabilities derived from massive pretraining. To counter this, we introduce CompressARC, a 76K parameter model without any pretraining that solves 20% of evaluation puzzles by minimizing the description length (MDL) of the target puzzle purely during inference time. The MDL endows CompressARC with extreme generalization abilities typically unheard of in deep learning. To our knowledge, CompressARC is the only deep learning method for ARC-AGI where training happens only on a single sample: the target inference puzzle itself, with the final solution information removed. Moreover, CompressARC does not train on the pre-provided ARC-AGI "training set". Under these extremely data-limited conditions, we do not ordinarily expect any puzzles to be solvable at all. Yet CompressARC still solves a diverse distribution of creative ARC-AGI puzzles, suggesting MDL to be an alternative feasible way to produce intelligence, besides conventional pretraining.

📄 PDF Abstract BibTeX arXiv:2512.06104

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

PASS: An ImageNet replacement for self-supervised pretraining without humans

2021-09-27 · NeurIPS Workshop ImageNet_PPF 2021 12 · Yuki M. Asano, Christian Rupprecht, Andrew Zisserman, Andrea Vedaldi

Computer vision has long relied on ImageNet and other large datasets of images sampled from the Internet for pretraining models. However, these datasets have ethical and technical shortcomings, such as containing persona…

BenchmarkingEthicsPose EstimationTransfer Learning

Pretraining Without Attention

2022-12-20 · Junxiong Wang, Jing Nathan Yan, Albert Gu, Alexander M. Rush

Transformers have been essential to pretraining success in NLP. While other architectures have been used, downstream accuracy is either significantly worse, or requires attention layers to match standard benchmarks such …

State Space Models

Revisiting Pretraining with Adapters

2021-08-01 · ACL (RepL4NLP) 2021 8 · Seungwon Kim, Alex Shum, Nathan Susanj, Jonathan Hilgart

Pretrained language models have served as the backbone for many state-of-the-art NLP results. These models are large and expensive to train. Recent work suggests that continued pretraining on task-specific data is worth …

Continual PretrainingTransfer Learning

Self-Pretraining for Small Datasets by Exploiting Patch Information

2021-01-01 · Zhang Chunyang

Deep learning tasks with small datasets are often tackled by pretraining models with large datasets on relevent tasks. Although pretraining methods mitigate the problem of overfitting, it can be difficult to find appro…

Strategies for Pretraining Neural Operators

2024-06-12 · Anthony Zhou, Cooper Lorsung, AmirPouya Hemmasian, Amir Barati Farimani

Pretraining for partial differential equation (PDE) modeling has recently shown promise in scaling neural operators across datasets to improve generalizability and performance. Despite these advances, our understanding o…

Transfer Learning