no code implementations • 14 Nov 2023 • Melanie Mitchell, Alessandro B. Palmarini, Arseny Moskvichev
We explore the abstract reasoning abilities of text-only and multimodal versions of GPT-4, using the ConceptARC benchmark [10], which is designed to evaluate robust understanding and reasoning with core-knowledge concepts.
1 code implementation • 23 May 2023 • Arseny Moskvichev, Ky-Vinh Mai
We show that our questions 1) adequately represent the source material 2) can be used to diagnose a model's memory capacity 3) are not trivial for modern language models even when the memory demand does not exceed those models' context lengths.
1 code implementation • 11 May 2023 • Arseny Moskvichev, Victor Vikram Odouard, Melanie Mitchell
In this paper we describe an in-depth evaluation benchmark for the Abstraction and Reasoning Corpus (ARC), a collection of few-shot abstraction and analogy problems developed by Chollet [2019].
no code implementations • 12 Apr 2021 • Arseny Moskvichev, James A. Liu
In this paper, we propose a novel transformer-based Updater-Extractor architecture and a training procedure that can work with sequences of arbitrary length and refine its knowledge about the world based on linguistic inputs.
1 code implementation • ICML Workshop LaReL 2020 • Marina Dubova, Arseny Moskvichev, Robert Goldstone
We examined the effects of social network organization on the properties of communication systems emerging in decentralized, multi-agent reinforcement learning communities.