Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Che, Liwei, Xue, Zhiyu, Quan, Yihao, Liu, Benlin, Shi, Zeru, Hurst, Michelle, Feldman, Jacob, Tang, Ruixiang, Krishna, Ranjay, Pavlovic, Vladimir
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914408747237376
author Che, Liwei
Xue, Zhiyu
Quan, Yihao
Liu, Benlin
Shi, Zeru
Hurst, Michelle
Feldman, Jacob
Tang, Ruixiang
Krishna, Ranjay
Pavlovic, Vladimir
author_facet Che, Liwei
Xue, Zhiyu
Quan, Yihao
Liu, Benlin
Shi, Zeru
Hurst, Michelle
Feldman, Jacob
Tang, Ruixiang
Krishna, Ranjay
Pavlovic, Vladimir
contents Counting serves as a simple but powerful test of a Large Vision-Language Model's (LVLM's) reasoning; it forces the model to identify each individual object and then add them all up. In this study, we investigate how LVLMs implement counting using controlled synthetic and real-world benchmarks, combined with mechanistic analyses. Our results show that LVLMs display a human-like counting behavior, with precise performance on small numerosities and noisy estimation for larger quantities. We introduce two novel interpretability methods, Visual Activation Patching and HeadLens, and use them to uncover a structured "counting circuit" that is largely shared across a variety of visual reasoning tasks. Building on these insights, we propose a lightweight intervention strategy that exploits simple and abundantly available synthetic images to fine-tune arbitrary pretrained LVLMs exclusively on counting. Despite the narrow scope of this fine-tuning, the intervention not only enhances counting accuracy on in-distribution synthetic data, but also yields an average improvement of +8.36% on out-of-distribution counting benchmarks and an average gain of +1.54% on complex, general visual reasoning tasks for Qwen2.5-VL. These findings highlight the central, influential role of counting in visual reasoning and suggest a potential pathway for improving overall visual reasoning capabilities through targeted enhancement of counting mechanisms.
format Preprint
id arxiv_https___arxiv_org_abs_2603_18523
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Models
Che, Liwei
Xue, Zhiyu
Quan, Yihao
Liu, Benlin
Shi, Zeru
Hurst, Michelle
Feldman, Jacob
Tang, Ruixiang
Krishna, Ranjay
Pavlovic, Vladimir
Computer Vision and Pattern Recognition
Artificial Intelligence
Counting serves as a simple but powerful test of a Large Vision-Language Model's (LVLM's) reasoning; it forces the model to identify each individual object and then add them all up. In this study, we investigate how LVLMs implement counting using controlled synthetic and real-world benchmarks, combined with mechanistic analyses. Our results show that LVLMs display a human-like counting behavior, with precise performance on small numerosities and noisy estimation for larger quantities. We introduce two novel interpretability methods, Visual Activation Patching and HeadLens, and use them to uncover a structured "counting circuit" that is largely shared across a variety of visual reasoning tasks. Building on these insights, we propose a lightweight intervention strategy that exploits simple and abundantly available synthetic images to fine-tune arbitrary pretrained LVLMs exclusively on counting. Despite the narrow scope of this fine-tuning, the intervention not only enhances counting accuracy on in-distribution synthetic data, but also yields an average improvement of +8.36% on out-of-distribution counting benchmarks and an average gain of +1.54% on complex, general visual reasoning tasks for Qwen2.5-VL. These findings highlight the central, influential role of counting in visual reasoning and suggest a potential pathway for improving overall visual reasoning capabilities through targeted enhancement of counting mechanisms.
title Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2603.18523