Search papers, labs, and topics across Lattice.
The paper investigates whether failures of Vision-Language Models (VLMs) on abstract visual reasoning tasks like Bongard problems stem from representational limitations or reasoning deficits. By converting visual inputs into symbolic representations derived from ground-truth generative programs, the authors isolate the reasoning capabilities of LLMs. Results on the Bongard-LOGO benchmark show that LLMs achieve significantly higher accuracy with symbolic inputs compared to VLMs processing raw images, indicating that visual representation is a major bottleneck.
VLMs' struggles with abstract visual reasoning aren't primarily due to weak reasoning, but rather a representational bottleneck in extracting the right symbolic information from pixels.
Vision--language models (VLMs) often fail on abstract visual reasoning benchmarks such as Bongard problems, raising the question of whether the main bottleneck lies in reasoning or representation. We study this on Bongard-LOGO, a synthetic benchmark of abstract concept learning with ground-truth generative programs, by comparing end-to-end VLMs on raw images with large language models (LLMs) given symbolic inputs derived from those images. Using symbolic inputs as a diagnostic probe rather than a practical multimodal architecture, our \emph{Componential--Grammatical (C--G)} paradigm reformulates Bongard-LOGO as a symbolic reasoning task based on LOGO-style action programs or structured descriptions. LLMs achieve large and consistent gains, reaching mid--90s accuracy on Free-form problems, while a strong visual baseline remains near chance under matched task definitions. Ablations on input format, explicit concept prompts, and minimal visual grounding show that these factors matter much less than the shift from pixels to symbolic structure. These results identify representation as a key bottleneck in abstract visual reasoning and show how symbolic input can serve as a controlled diagnostic upper bound.