Hierarchical Process Reward Models are Symbolic Vision Learners

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Shan, Chen, Aotian, Zou, Kai, Gu, Jindong, Xue, Yuan, Hengel, Anton van den
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915650717351936
author Zhang, Shan
Chen, Aotian
Zou, Kai
Gu, Jindong
Xue, Yuan
Hengel, Anton van den
author_facet Zhang, Shan
Chen, Aotian
Zou, Kai
Gu, Jindong
Xue, Yuan
Hengel, Anton van den
contents Symbolic computer vision represents diagrams through explicit logical rules and structured representations, enabling interpretable understanding in machine vision. This requires fundamentally different learning paradigms from pixel-based visual models. Symbolic visual learners parse diagrams into geometric primitives-points, lines, and shapes-whereas pixel-based learners operate on textures and colors. We propose a novel self-supervised symbolic auto-encoder that encodes diagrams into structured primitives and their interrelationships within the latent space, and decodes them through our executable engine to reconstruct the input diagrams. Central to this architecture is Symbolic Hierarchical Process Reward Modeling, which applies hierarchical step-level parsing rewards to enforce point-on-line, line-on-shape, and shape-on-relation consistency. Since vanilla reinforcement learning exhibits poor exploration in the policy space during diagram reconstruction; we thus introduce stabilization mechanisms to balance exploration and exploitation. We fine-tune our symbolic encoder on downstream tasks, developing a neuro-symbolic system that integrates the reasoning capabilities of neural networks with the interpretability of symbolic models through reasoning-grounded visual rewards. Evaluations across reconstruction, perception, and reasoning tasks demonstrate the effectiveness of our approach: achieving a 98.2% reduction in MSE for geometric diagram reconstruction, surpassing GPT-4o by 0.6% with a 7B model on chart reconstruction, and improving by +13% on the MathGlance perception benchmark, and by +3% on MathVerse and GeoQA reasoning benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2512_03126
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Hierarchical Process Reward Models are Symbolic Vision Learners
Zhang, Shan
Chen, Aotian
Zou, Kai
Gu, Jindong
Xue, Yuan
Hengel, Anton van den
Computer Vision and Pattern Recognition
Symbolic computer vision represents diagrams through explicit logical rules and structured representations, enabling interpretable understanding in machine vision. This requires fundamentally different learning paradigms from pixel-based visual models. Symbolic visual learners parse diagrams into geometric primitives-points, lines, and shapes-whereas pixel-based learners operate on textures and colors. We propose a novel self-supervised symbolic auto-encoder that encodes diagrams into structured primitives and their interrelationships within the latent space, and decodes them through our executable engine to reconstruct the input diagrams. Central to this architecture is Symbolic Hierarchical Process Reward Modeling, which applies hierarchical step-level parsing rewards to enforce point-on-line, line-on-shape, and shape-on-relation consistency. Since vanilla reinforcement learning exhibits poor exploration in the policy space during diagram reconstruction; we thus introduce stabilization mechanisms to balance exploration and exploitation. We fine-tune our symbolic encoder on downstream tasks, developing a neuro-symbolic system that integrates the reasoning capabilities of neural networks with the interpretability of symbolic models through reasoning-grounded visual rewards. Evaluations across reconstruction, perception, and reasoning tasks demonstrate the effectiveness of our approach: achieving a 98.2% reduction in MSE for geometric diagram reconstruction, surpassing GPT-4o by 0.6% with a 7B model on chart reconstruction, and improving by +13% on the MathGlance perception benchmark, and by +3% on MathVerse and GeoQA reasoning benchmarks.
title Hierarchical Process Reward Models are Symbolic Vision Learners
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.03126