TraceViT: Grounded Trace Supervision for Visual Abstract Reasoning
cs.CV, cs.AI
Submitted: 2026-07-31
Updated: 2026-11-05
Code: https://github.com/LiuBinnan/TraceViT
Project page: https://lambdalabsml.github.io/ARC2025_
Terminology
Sources
- On the Measure of Intelligence
- ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems
- Don't throw the baby out with the bathwater: How and why deep learning for ARC
- Slots, Transitions, Loops: Learning Composable World Models for ARC
- Addressing the Abstraction and Reasoning Corpus via Procedural Example Generation
- ARC Is a Vision Problem!
- Less is More: Recursive Reasoning with Tiny Networks
- ARC-TGI: Human-Validated Task Generators with Reasoning Chain Templates for ARC-AGI
- RoSeViT: Role-Separated Vision Transformers for ARC Visual Reasoning
- Test-time Adaptation of Tiny Recursive Models
- ARC-GEN: A Mimetic Procedural Benchmark Generator for the Abstraction and Reasoning Corpus
- Towards Efficient Neurally-Guided Program Induction for ARC-AGI
- Tiny Recursive Models on ARC-AGI-1: Inductive Biases, Identity Conditioning, and Test-Time Compute
- LoopViT: Scaling Visual ARC with Looped Transformers
- Hierarchical Reasoning Model
- Tiny Recursive Reasoning with Mamba-2 Attention Hybrid
- Think Visually, Reason Textually: Vision-Language Synergy in ARC
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs