Large language models as general pattern machines
We observe that pre-trained large language models (LLMs) are capable of autoregressively
completing complex token sequences--from arbitrary ones procedurally generated by …
completing complex token sequences--from arbitrary ones procedurally generated by …
Hypothesis search: Inductive reasoning with language models
Inductive reasoning is a core problem-solving capacity: humans can identify underlying
principles from a few examples, which can then be robustly generalized to novel scenarios …
principles from a few examples, which can then be robustly generalized to novel scenarios …
The conceptarc benchmark: Evaluating understanding and generalization in the arc domain
The abilities to form and abstract concepts is key to human intelligence, but such abilities
remain lacking in state-of-the-art AI systems. There has been substantial research on …
remain lacking in state-of-the-art AI systems. There has been substantial research on …
Llms and the abstraction and reasoning corpus: Successes, failures, and the importance of object-based representations
Can a Large Language Model (LLM) solve simple abstract reasoning problems? We explore
this broad question through a systematic analysis of GPT on the Abstraction and Reasoning …
this broad question through a systematic analysis of GPT on the Abstraction and Reasoning …
ARCLE: The Abstraction and Reasoning Corpus Learning Environment for Reinforcement Learning
This paper introduces ARCLE, an environment designed to facilitate reinforcement learning
research on the Abstraction and Reasoning Corpus (ARC). Addressing this inductive …
research on the Abstraction and Reasoning Corpus (ARC). Addressing this inductive …
Abstract visual reasoning enabled by language
G Camposampiero, L Houmard… - Proceedings of the …, 2023 - openaccess.thecvf.com
While artificial intelligence (AI) models have achieved human or even superhuman
performance in many well-defined applications, they still struggle to show signs of broad and …
performance in many well-defined applications, they still struggle to show signs of broad and …
Generalized planning for the abstraction and reasoning corpus
The Abstraction and Reasoning Corpus (ARC) is a general artificial intelligence benchmark
that poses difficulties for pure machine learning methods due to its requirement for fluid …
that poses difficulties for pure machine learning methods due to its requirement for fluid …
Unraveling the arc puzzle: Mimicking human solutions with object-centric decision transformer
In the pursuit of artificial general intelligence (AGI), we tackle Abstraction and Reasoning
Corpus (ARC) tasks using a novel two-pronged approach. We employ the Decision …
Corpus (ARC) tasks using a novel two-pronged approach. We employ the Decision …
Relational decomposition for program synthesis
C Hocquette, A Cropper - arXiv preprint arXiv:2408.12212, 2024 - arxiv.org
We introduce a novel approach to program synthesis that decomposes complex functional
tasks into simpler relational synthesis sub-tasks. We demonstrate the effectiveness of our …
tasks into simpler relational synthesis sub-tasks. We demonstrate the effectiveness of our …
Neural networks for abstraction and reasoning: Towards broad generalization in machines
M Bober-Irizar, S Banerjee - arXiv preprint arXiv:2402.03507, 2024 - arxiv.org
For half a century, artificial intelligence research has attempted to reproduce the human
qualities of abstraction and reasoning-creating computer systems that can learn new …
qualities of abstraction and reasoning-creating computer systems that can learn new …