Copertina

Do LLMs Build Spatial World Models? Evidence from Grid-World Maze Tasks

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Li, Weijiang, Zhu, Yilin, Das, Rajarshi, Dube, Parijat
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Do LLMs Build World Models From Text? A Multilingual Diagnostic of Spatial Reasoning
di: Pan, Zhikai, et al.
Pubblicazione: (2026)

Transformers Use Causal World Models in Maze-Solving Tasks
di: Spies, Alex F., et al.
Pubblicazione: (2024)

From Text to Space: Mapping Abstract Spatial Models in LLMs during a Grid-World Navigation Task
di: Martorell, Nicolas
Pubblicazione: (2025)

OnlySportsLM: Optimizing Sports-Domain Language Models with SOTA Performance under Billion Parameters
di: Chen, Zexin, et al.
Pubblicazione: (2024)

Thinking by Doing: Building Efficient World Model Reasoning in LLMs via Multi-turn Interaction
di: Shu, Bao, et al.
Pubblicazione: (2025)

Test Time Learning for Time Series Forecasting
di: Christou, Panayiotis, et al.
Pubblicazione: (2024)

Natural Building Blocks for Structured World Models: Theory, Evidence, and Scaling
di: Da Costa, Lancelot, et al.
Pubblicazione: (2025)

Moral Mazes in the Era of LLMs
di: Nguyen, Dang, et al.
Pubblicazione: (2026)

Impact of Noise on LLM-Models Performance in Abstraction and Reasoning Corpus (ARC) Tasks with Model Temperature Considerations
di: Khandalkar, Nikhil, et al.
Pubblicazione: (2025)

Reinforcing VLAs in Task-Agnostic World Models
di: Wang, Yucen, et al.
Pubblicazione: (2026)

Code-Driven Planning in Grid Worlds with Large Language Models
di: Aravindan, Ashwath Vaithinathan, et al.
Pubblicazione: (2025)

What Do LLM Agents Know About Their World? Task2Quiz: A Paradigm for Studying Environment Understanding
di: Liu, Siyuan, et al.
Pubblicazione: (2026)

TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks
di: Chu, Zhaoyang, et al.
Pubblicazione: (2026)

Embodied AI: From LLMs to World Models
di: Feng, Tongtong, et al.
Pubblicazione: (2025)

The Labyrinth of Links: Navigating the Associative Maze of Multi-modal LLMs
di: Li, Hong, et al.
Pubblicazione: (2024)

Cross-Cultural Transfer of Commonsense Reasoning in LLMs: Evidence from the Arab World
di: Almheiri, Saeed, et al.
Pubblicazione: (2025)

World2Mind: Cognition Toolkit for Allocentric Spatial Reasoning in Foundation Models
di: Ruan, Shouwei, et al.
Pubblicazione: (2026)

Vector Quantization in the Brain: Grid-like Codes in World Models
di: Peng, Xiangyuan, et al.
Pubblicazione: (2025)

Linear Spatial World Models Emerge in Large Language Models
di: Tehenan, Matthieu, et al.
Pubblicazione: (2025)

Multi-Task Interactive Robot Fleet Learning with Visual World Models
di: Liu, Huihan, et al.
Pubblicazione: (2024)

Exploring Next Token Prediction in Theory of Mind (ToM) Tasks: Comparative Experiments with GPT-2 and LLaMA-2 AI Models
di: Yadav, Pavan, et al.
Pubblicazione: (2025)

SMDD-Bench: Can LLMs Solve Real-World Small Molecule Drug Design Tasks?
di: Han, Kevin, et al.
Pubblicazione: (2026)

WorldCoder, a Model-Based LLM Agent: Building World Models by Writing Code and Interacting with the Environment
di: Tang, Hao, et al.
Pubblicazione: (2024)

Task Expansion and Cross Refinement for Open-World Conditional Modeling
di: Brahmavar, Shreyas Bhat, et al.
Pubblicazione: (2026)

Affordances Enable Partial World Modeling with LLMs
di: Khetarpal, Khimya, et al.
Pubblicazione: (2026)

MobileDreamer: Generative Sketch World Model for GUI Agent
di: Cao, Yilin, et al.
Pubblicazione: (2026)

Do Multi-Agents Dream of Electric Screens? Achieving Perfect Accuracy on AndroidWorld Through Task Decomposition
di: Favreau, Pierre-Louis, et al.
Pubblicazione: (2026)

Embodied World Models Emerge from Navigational Task in Open-Ended Environments
di: Jin, Li, et al.
Pubblicazione: (2025)

Back to Parsimonious Latents: Learning Task-Centric World Models from Visual Foundations
di: Fu, Minghao, et al.
Pubblicazione: (2026)

Scalable Task Planning via Large Language Models and Structured World Representations
di: Pérez-Dattari, Rodrigo, et al.
Pubblicazione: (2024)

PlotMap: Automated Layout Design for Building Game Worlds
di: Wang, Yi, et al.
Pubblicazione: (2023)

Do Transformer World Models Give Better Policy Gradients?
di: Ma, Michel, et al.
Pubblicazione: (2024)

Lumine: An Open Recipe for Building Generalist Agents in 3D Open Worlds
di: Tan, Weihao, et al.
Pubblicazione: (2025)

From Stateless to Situated: Building a Psychological World for LLM-Based Emotional Support
di: Zhao, Boning, et al.
Pubblicazione: (2026)

DexWorldModel: Causal Latent World Modeling towards Automated Learning of Embodied Tasks
di: Deng, Yueci, et al.
Pubblicazione: (2026)

The World is Not Mono: Enabling Spatial Understanding in Large Audio-Language Models
di: You, Yuhuan, et al.
Pubblicazione: (2026)

Towards Automated Fact-Checking of Real-World Claims: Exploring Task Formulation and Assessment with LLMs
di: Sahitaj, Premtim, et al.
Pubblicazione: (2025)

PWM: Policy Learning with Multi-Task World Models
di: Georgiev, Ignat, et al.
Pubblicazione: (2024)

Convergent World Representations and Divergent Tasks
di: Park, Core Francisco
Pubblicazione: (2026)

WorldGPT: Empowering LLM as Multimodal World Model
di: Ge, Zhiqi, et al.
Pubblicazione: (2024)