REVIEW 7 cited by
LEGO-Puzzles: How Good Are MLLMs at Multi-Step Spatial Reasoning?
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Multi-step spatial reasoning entails understanding and reasoning about spatial relationships across multiple sequential steps, which is crucial for tackling complex real-world applications, such as robotic manipulation, autonomous navigation, and automated assembly. To assess how well current Multimodal Large Language Models (MLLMs) have acquired this fundamental capability, we introduce LEGO-Puzzles, a scalable benchmark designed to evaluate both spatial understanding and sequential reasoning in MLLMs through LEGO-based tasks. LEGO-Puzzles consists of 1,100 carefully curated visual question-answering (VQA) samples spanning 11 distinct tasks, ranging from basic spatial understanding to complex multi-step reasoning. Based on LEGO-Puzzles, we conduct a comprehensive evaluation of 20 state-of-the-art MLLMs and uncover significant limitations in their spatial reasoning capabilities: even the most powerful MLLMs can answer only about half of the test cases, whereas human participants achieve over 90% accuracy. Furthermore, based on LEGO-Puzzles, we design generation tasks to investigate whether MLLMs can transfer their spatial understanding and reasoning abilities to image generation. Our experiments show that only GPT-4o and Gemini-2.0-Flash exhibit a limited ability to follow these instructions, while other MLLMs either replicate the input image or generate completely irrelevant outputs. Overall, LEGO-Puzzles exposes critical deficiencies in existing MLLMs' spatial understanding and sequential reasoning capabilities, and underscores the need for further advancements in multimodal spatial reasoning.
Forward citations
Cited by 7 Pith papers
-
SpaceVista: All-Scale Visual Spatial Reasoning from mm to km
SpaceVista contributes a 1M-QA, 38K-video all-scale spatial reasoning dataset spanning mm to km, a manually verified benchmark, and a fine-tuned 7B MLLM with scale experts and progressive reward training.
-
Diagnosing Corruption-Induced Reliability Failures in Vision-Language Models
Mild visual corruption can boost a vision-language model's top-1 accuracy while its confidence–correctness alignment (measured by the new RAS score) degrades.
-
SpatialThinker: Reinforcing Scene Graph-Grounded Spatial Reasoning via Dense Rewards
Dense scene-graph-grounded rewards let a 7B multimodal LLM trained on 7K synthetic questions beat SFT and sparse-RL baselines and outscore GPT-4o on average across 12 spatial/real-world benchmarks.
-
CharacterShot: Controllable and Consistent 4D Character Animation
A new pipeline generates pose-controlled, view-consistent 4D character animations from one reference image and a 2D pose sequence, backed by a new 13,115-character dataset and benchmark.
-
Ascending the Infinite Ladder: Benchmarking Spatial Deformation Reasoning in Vision-Language Models
Vision-language models can handle some 2D shape puzzles but nearly all fail at multi-step 3D spatial deformation reasoning.
-
Jigsaw-Puzzles: From Seeing to Understanding to Reasoning in Vision-Language Models
Current vision-language models fall far short of humans on spatial reasoning, especially when they must generate answers directly instead of choosing from options.
-
STAR-R1: Spatial TrAnsformation Reasoning by Reinforcing Multimodal LLMs
STAR-R1 uses single-stage reinforcement learning with fine-grained rewards to improve spatial transformation reasoning in multimodal LLMs, outperforming supervised fine-tuning on cross-view TVR tasks.
Discussion (0). Continue with ORCID to comment.