ARC-AGI-2 Technical Report

• URL: http://arxiv.org/abs/2603.06590v1
• Date: Wed, 04 Feb 2026 10:03:56 GMT
• Title: ARC-AGI-2 Technical Report
• Abstract summary: Abstraction and Reasoning Corpus (ARC) is designed to assess generalization beyond pattern matching.<n>We present a transformer-based system that advances ARC performance by combining neural inference with structure-aware priors and online task adaptation.
• Score: 0.0846998909761807
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: The Abstraction and Reasoning Corpus (ARC) is designed to assess generalization beyond pattern matching, requiring models to infer symbolic rules from very few examples. In this work, we present a transformer-based system that advances ARC performance by combining neural inference with structure-aware priors and online task adaptation. Our approach is built on four key ideas. First, we reformulate ARC reasoning as a sequence modeling problem using a compact task encoding with only 125 tokens, enabling efficient long-context processing with a modified LongT5 architecture. Second, we introduce a principled augmentation framework based on group symmetries, grid traversals, and automata perturbations, enforcing invariance to representation changes. Third, we apply test-time training (TTT) with lightweight LoRA adaptation, allowing the model to specialize to each unseen task by learning its transformation logic from demonstrations. Fourth, we design a symmetry-aware decoding and scoring pipeline that aggregates likelihoods across augmented task views, effectively performing ``multi-perspective reasoning'' over candidate solutions. We demonstrate that these components work synergistically: augmentations expand hypothesis space, TTT sharpens local reasoning, and symmetry-based scoring improves solution consistency. Our final system achieves a significant improvement over transformer baselines and surpasses prior neural ARC solvers, closing the gap toward human-level generalization.

Related papers

• TMallGS: Scaling Unified Feature and Sequence Modeling for Generative E-commerce Search [4.061689397024572]
  We propose TmallGS, a scalable ranking architecture for Tmall search.<n>TmallGS includes five key components: (1) Hierarchical Distribution-Calibrated Tokenization; (2) a Field-Adaptive Gated Transformer Backbone; (3) Decoupled FiLM Late Fusion to preserve explicit high-frequency signals; and (4) a Context-Aware Bias Net.<n>Extensive offline experiments and online A/B tests on Tmall Search show that TmallGS improves training throughput and achieves substantial gains in UCTCVR and GMV.
  arXiv  Detail & Related papers  (2026-07-15T02:52:22Z)
• The Art of Interrogation: Consistency Amplifies Factuality in Spatial Reasoning [65.67266333751569]
  Current Large Reasoning Models (LRMs) exhibit remarkable general capabilities but significantly underperform in spatial reasoning tasks.<n>We propose a self-supervised reinforcement learning framework that targets the internal reasoning process without requiring ground-truth annotations.<n>We show that this label-free consistency training approaches the accuracy of models trained with ground-truth supervision and achieves similar generalization across diverse tasks and data domains.
  arXiv  Detail & Related papers  (2026-06-10T10:50:06Z)
• Compositional Neuro-Symbolic Reasoning [0.7894152109655374]
  We study structured abstraction-based reasoning for the Abstraction and Reasoning Corpus (ARC)<n>We propose a neuro-symbolic architecture that extracts object-level structure from grids.<n>We open-source the ARC-AGI-2 Reasoner code.
  arXiv  Detail & Related papers  (2026-04-02T18:04:34Z)
• Dual-Phase LLM Reasoning: Self-Evolved Mathematical Frameworks [48.105258051884384]
  This paper proposes a new two-stage training framework that enhances models' self-correction capabilities.<n>During the first stage, a multi-turn dialogue strategy guides the model to generate long chain-of-thought (CoT) data.<n>The second stage employs a difficulty-aware rejection sampling mechanism to dynamically optimize data distribution.
  arXiv  Detail & Related papers  (2026-01-09T08:19:11Z)
• CSMCIR: CoT-Enhanced Symmetric Alignment with Memory Bank for Composed Image Retrieval [54.15776146365823]
  Composed Image Retrieval (CIR) enables users to search for target images using both a reference image and manipulation text.<n>We propose CSMCIR, a unified representation framework that achieves efficient query-target alignment through three synergistic components.
  arXiv  Detail & Related papers  (2026-01-07T09:21:38Z)
• Unified token representations for sequential decision models [0.07812854697536452]
  We propose a Unified Token Representation (UTR) that merges return-to-go, state, and action into a single token, substantially reducing sequence length and model complexity.<n>We develop two variants: UDT and UDC, built upon transformer and gated CNN backbones, respectively. Both achieve comparable or superior performance to state-of-the-art methods with markedly lower computation.
  arXiv  Detail & Related papers  (2025-10-24T13:25:43Z)
• BERT4beam: Large AI Model Enabled Generalized Beamforming Optimization [77.17508487745026]
  This paper investigates the large-scale AI model designed for beamforming optimization to adapt and generalize to diverse tasks defined by system utilities and scales.<n>We propose a novel framework based on bidirectional encoder representations from transformers (BERT), termed BERT4 encoder.<n>Based on the framework, we propose two BERT-based approaches for single-task and multi-task beamforming optimization, respectively.
  arXiv  Detail & Related papers  (2025-09-14T02:49:29Z)
• Small transformer architectures for task switching [2.7195102129095003]
  It is non-trivial to conceive small-scale applications for which attention-based architectures outperform traditional approaches.<n>We show that standard transformers cannot solve a basic task switching reference model.<n>We show that transformers, long short-term memory recurrent networks (LSTM), and plain multi-layer perceptrons (MLPs) achieve similar, but only modest prediction accuracies.
  arXiv  Detail & Related papers  (2025-08-06T14:01:05Z)
• LARES: Latent Reasoning for Sequential Recommendation [96.26996622771593]
  We present LARES, a novel and scalable LAtent REasoning framework for Sequential recommendation.<n>Our proposed approach employs a recurrent architecture that allows flexible expansion of reasoning depth without increasing parameter complexity.<n>Our framework exhibits seamless compatibility with existing advanced models, further improving their recommendation performance.
  arXiv  Detail & Related papers  (2025-05-22T16:22:54Z)
• End-to-End Meta-Bayesian Optimisation with Transformer Neural Processes [52.818579746354665]
  This paper proposes the first end-to-end differentiable meta-BO framework that generalises neural processes to learn acquisition functions via transformer architectures. We enable this end-to-end framework with reinforcement learning (RL) to tackle the lack of labelled acquisition data.
  arXiv  Detail & Related papers  (2023-05-25T10:58:46Z)

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.