# WaferSAGE: Large Language Model-Powered Wafer Defect Analysis via Synthetic Data Generation and Rubric-Guided Reinforcement Learning

arXiv: https://arxiv.org/abs/2604.27629 · HTML: https://mlsanity.com/paper/2604.27629/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Chart understanding and anomaly detection, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 20 of its 34 references that are in the map.
- First posted on arXiv 30 Apr 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Chart understanding and anomaly detection](https://mlsanity.com/mllm/chart-understanding-and-anomaly-detection/)
- Synthetic data and self-improvement for LLMs / [Self-improving MLLMs and synthetic data](https://mlsanity.com/synth/self-improving-mllms-and-synthetic-data/)

## Builds on (most cited first, 20 of 20)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models](https://mlsanity.com/paper/2304.10592/) (2023)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [ShareGPT4V: Improving Large Multi-Modal Models with Better Captions](https://mlsanity.com/paper/2311.12793/) (2023)
- [Group Sequence Policy Optimization](https://mlsanity.com/paper/2507.18071/) (2025)
- [WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation](https://mlsanity.com/paper/2303.14814/) (2023)
- [AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection](https://mlsanity.com/paper/2310.18961/) (2023)
- [IAD-R1: Reinforcing Consistent Reasoning in Industrial Anomaly Detection](https://mlsanity.com/paper/2508.09178/) (2025)
- [EMIT: Enhancing MLLMs for Industrial Anomaly Detection via Difficulty-Aware GRPO](https://mlsanity.com/paper/2507.21619/) (2025)
- [Omni-RRM: Advancing Omni Reward Modeling via Automatic Rubric-Grounded Preference Synthesis](https://mlsanity.com/paper/2602.00846/) (2026)
- [Wafer Map Defect Patterns Semi-Supervised Classification Using Latent Vector Representation](https://mlsanity.com/paper/2311.12840/) (2023)
- [Wafer2Spike: Spiking Neural Network for Wafer Map Pattern Classification](https://mlsanity.com/paper/2411.19422/) (2024)
- [Wafer Map Defect Classification Using Autoencoder-Based Data Augmentation and Convolutional Neural Network](https://mlsanity.com/paper/2411.11029/) (2024)
- [Smooth Operator: Smooth Verifiable Reward Activates Spatial Reasoning Ability of Vision-Language Model](https://mlsanity.com/paper/2601.07695/) (2026)
- [Semiconductor Wafer Map Defect Classification with Tiny Vision Transformers](https://mlsanity.com/paper/2504.02494/) (2025)

## Built on by (most cited first, 1 of 1)

- [From Holistic Evaluation to Structured Criteria: Rubrics Across the Evolving LLM Landscape](https://mlsanity.com/paper/2606.08625/) (2026)

## Abstract (arXiv)

We present WaferSAGE, a framework for wafer defect visual question answering using small vision-language models. To address data scarcity in semiconductor manufacturing, we propose a three-stage synthesis pipeline incorporating structured rubric generation for precise evaluation. Starting from limited labeled wafer maps, we employ clustering-based cleaning to filter label noise, then generate comprehensive defect descriptions using vision-language models, which are converted into structured evaluation rubrics criteria. These rubrics guide the synthesis of VQA pairs, ensuring coverage across defect type identification, spatial distribution, morphology, and root cause analysis. Our dual assessment framework aligns rule-based metrics with LLM-Judge scores via Bayesian optimization, enabling reliable automated evaluation. Through curriculum-based reinforcement learning with Group Sequence Policy Optimization (GSPO) and rubric-aligned rewards, our 4B-parameter Qwen3-VL model achieves a 6.493 LLM-Judge score, closely approaching Gemini-3-Flash (7.149) while enabling complete on-premise deployment. We demonstrate that small models with domain-specific training can surpass proprietary large models in specialized industrial visual understanding, offering a viable path for privacy-preserving, cost-effective deployment in semiconductor manufacturing.