Spatial Reasoning in LLM Game Agents: Impact of Causal Context and Multi-Step Planning

• URL: http://arxiv.org/abs/2607.22732v1
• Date: Wed, 22 Jul 2026 12:10:45 GMT
• Title: Spatial Reasoning in LLM Game Agents: Impact of Causal Context and Multi-Step Planning
• Abstract summary: We conduct experiments across varying model scales, reasoning modes, and planning horizons.<n>We introduce a focused GVGAI benchmark consisting of three custom games with five difficulty levels to isolate spatial navigation.<n>Our results show that while larger models with an enabled thinking mode identify their positions more accurately, overall performance in coordinate matching remains limited for smaller models.
• Score: 56.67877550139192
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: LLM-based game agents often perform poorly on more complex tasks. This work examines whether these failures are linked to limited spatial reasoning and evaluates whether causal prompt augmentation and multi-step planning can improve win-rates while managing response latency. Using the open-source Qwen3 model family, we conduct experiments across varying model scales, reasoning modes, and planning horizons. We further introduce a focused GVGAI benchmark consisting of three custom games with five difficulty levels to isolate spatial navigation. The evaluation follows two paradigms: an initial ``positioning experiment'' to test an agent's ability to find its exact coordinates, and a study of game-play success. Our results show that while larger models with an enabled thinking mode identify their positions more accurately, overall performance in coordinate matching remains limited for smaller models. Win rates decrease as game levels and layout complexity increase, validating the benchmark's difficulty scaling. Integrating causal context into the prompts tends to improve the agents' success rates, particularly for bigger models. While enabling thinking mode and longer planning horizons significantly improve performance, multi-step planning further reduces mean per-step response times, offering a practical trade-off between reasoning depth and execution speed.

Related papers

• WISE: A Long-Horizon Agent in Minecraft with Why-Which Reasoning [6.545488752368577]
  We propose WISE, a long-horizon agent framework with an enhanced low-level controller equipped with a Causal Event Graph.<n>WISE augments episodic memory with explicit causal structure linking observations to task relevance.<n>Experiments show that WISE largely improves task success and efficiency on long-horizon sparse tasks.
  arXiv  Detail & Related papers  (2026-06-11T03:35:02Z)
• MINDGAMES: A Live Arena for Evaluating Social and Strategic Reasoning in Multi-Agent LLMs [54.81359054218573]
  We introduce Mindgames, a multi-game arena and evaluation platform for large language models (LLMs)<n>Mindgames provides a unified interaction interface, TrueSkill-based rating, and full trajectory logging across four game environments.<n>We analyze 29,571 multi-agent games with turn-level observations, actions, and rewards, together with MG-Ref, a deterministic offline tournament protocol.
  arXiv  Detail & Related papers  (2026-05-28T07:33:47Z)
• AgentCollab: A Self-Evaluation-Driven Collaboration Paradigm for Efficient LLM Agents [37.232397795331444]
  We present AgentCollab, a self-driven collaborative inference framework that dynamically coordinates models with different reasoning capacities during agent execution.<n>Instead of relying on external routing modules, the framework uses the agent's own self-reflection signal to determine whether the current reasoning trajectory is making meaningful progress.
  arXiv  Detail & Related papers  (2026-03-27T03:07:34Z)
• LLM CHESS: Benchmarking Reasoning and Instruction-Following in LLMs through Chess [30.797553771114746]
  We introduce LLM CHESS, an evaluation framework designed to probe the generalization of reasoning and instruction-following abilities in large language models (LLMs)<n>We rank over 50 open and closed source models by playing against a random opponent using a range of behavioral metrics, including move quality, move legality, hallucinated actions, and game duration.<n>For a subset of top reasoning models, we derive an Elo estimate by playing against a chess engine with variably configured skill, which allows for comparisons between models in an easily understandable way.
  arXiv  Detail & Related papers  (2025-12-01T18:51:08Z)
• Staying in the Sweet Spot: Responsive Reasoning Evolution via Capability-Adaptive Hint Scaffolding [59.60915947702282]
  Reinforcement learning with verifiable rewards (RLVR) has achieved remarkable success in enhancing the reasoning capabilities of large language models (LLMs)<n>Existing RLVR methods often suffer from exploration inefficiency due to mismatches between the training data's difficulty and the model's capability.<n>We propose SEELE, a novel supervision-aided RLVR framework that dynamically adjusts problem difficulty to stay within the high-efficiency region.
  arXiv  Detail & Related papers  (2025-09-08T17:36:21Z)
• CausalPlan: Empowering Efficient LLM Multi-Agent Collaboration Through Causality-Driven Planning [25.322580535468013]
  CausalPlan is a framework that integrates explicit structural causal reasoning into the large language model (LLM) planning process.<n>We evaluate CausalPlan on the Overcooked-AI benchmark across five multi-agent coordination tasks and four LLMs of varying sizes.<n>Results show that CausalPlan consistently reduces invalid actions and improves collaboration in both AI-AI and human-AI settings.
  arXiv  Detail & Related papers  (2025-08-19T10:37:20Z)
• MetaScale: Test-Time Scaling with Evolving Meta-Thoughts [51.35594569020857]
  Experimental results demonstrate that MetaScale consistently outperforms standard inference approaches.<n> METASCALE scales more effectively with increasing sampling budgets and produces more structured, expert-level responses.
  arXiv  Detail & Related papers  (2025-03-17T17:59:54Z)
• TMGBench: A Systematic Game Benchmark for Evaluating Strategic Reasoning Abilities of LLMs [45.12542636218608]
  We propose TMGBench, characterized by comprehensive game type coverage, diverse scenarios and flexible game organization.<n>Specifically, we incorporate all 144 game types summarized by the Robinson-Goforth topology of 2x2 games, constructed as classic games in our benchmark.<n>To provide a sustainable evaluation framework adaptable to increasingly powerful LLMs, we treat the aforementioned games as atomic units.
  arXiv  Detail & Related papers  (2024-10-14T13:15:34Z)
• Improving Long-Horizon Imitation Through Instruction Prediction [93.47416552953075]
  In this work, we explore the use of an often unused source of auxiliary supervision: language. Inspired by recent advances in transformer-based models, we train agents with an instruction prediction loss that encourages learning temporally extended representations that operate at a high level of abstraction. In further analysis we find that instruction modeling is most important for tasks that require complex reasoning, while understandably offering smaller gains in environments that require simple plans.
  arXiv  Detail & Related papers  (2023-06-21T20:47:23Z)
• Off-Beat Multi-Agent Reinforcement Learning [62.833358249873704]
  We investigate model-free multi-agent reinforcement learning (MARL) in environments where off-beat actions are prevalent. We propose a novel episodic memory, LeGEM, for model-free MARL algorithms. We evaluate LeGEM on various multi-agent scenarios with off-beat actions, including Stag-Hunter Game, Quarry Game, Afforestation Game, and StarCraft II micromanagement tasks.
  arXiv  Detail & Related papers  (2022-05-27T02:21:04Z)

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.