# PAIQ: Patch-Aligned Semantic Injection via Residual Rotation

arXiv: https://arxiv.org/abs/2609.37685 · HTML: https://mlsanity.com/paper/2609.37685/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 21 references that are in the map.
- First posted on arXiv 29 Sep 2026.

## Builds on (most cited first, 16 of 16)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distances](https://mlsanity.com/paper/1306.0895/) (2013)
- [SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension](https://mlsanity.com/paper/2307.16125/) (2023)
- [Attention is not Explanation](https://mlsanity.com/paper/1902.10186/) (2019)
- [A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge](https://mlsanity.com/paper/2206.01718/) (2022)
- [LLaVA-Mini: Efficient Image and Video Large Multimodal Models with One Vision Token](https://mlsanity.com/paper/2501.03895/) (2025)
- [Eagle: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders](https://mlsanity.com/paper/2408.15998/) (2024)
- [From CLIP to DINO: Visual Encoders Shout in Multi-modal Large Language Models](https://mlsanity.com/paper/2310.08825/) (2023)
- [DINOv2 Meets Text: A Unified Framework for Image- and Pixel-Level Vision-Language Alignment](https://mlsanity.com/paper/2412.16334/) (2024)
- [BRAVE: Broadening the visual encoding of vision-language models](https://mlsanity.com/paper/2404.07204/) (2024)
- [Cross-modal Information Flow in Multimodal Large Language Models](https://mlsanity.com/paper/2411.18620/) (2024)
- [FLUX-Reason-6M & PRISM-Bench: A Million-Scale Text-to-Image Reasoning Dataset and Comprehensive Benchmark](https://mlsanity.com/paper/2509.09680/) (2025)
- [Unifying Specialized Visual Encoders for Video Language Models](https://mlsanity.com/paper/2501.01426/) (2025)
- [Granulon: Awakening Pixel-Level Visual Encoders with Adaptive Multi-Granularity Semantics for MLLM](https://mlsanity.com/paper/2603.08800/) (2026)
- [CoME-VL: Scaling Complementary Multi-Encoder Vision-Language Learning](https://mlsanity.com/paper/2604.03231/) (2026)

## Abstract (arXiv)

Language-aligned and self-supervised visual encoders offer complementary strengths in semantic abstraction and spatial detail. Harnessing this complementarity requires enriching local features while retaining distinctions between semantically related patches. We introduce PAIQ, a patch-aligned semantic injection framework that combines content-based cross-encoder matching with orthogonally constrained residual updates. Using DINOv3 patch features as the spatial base, PAIQ aggregates complementary SigLIP features through joint source allocation and injects the aggregate--base differences through a shared orthogonal transformation Q. This rotation adapts update directions while preserving residual norms and pairwise angles. For fixed projected features, we derive conditions for patch separability under similar semantic aggregates and show that rotation adds a nonnegative separation term over direct interpolation when the aggregate is shared. Only the projection and fusion parameters are trained; both visual encoders and the language model remain frozen, and fusion retains 196 visual tokens. Across diverse language backbones, PAIQ yields broad gains in judge-assessed correctness and reductions in hallucination severity over single-encoder interfaces on image description and visual question answering. On the 2B and 9B Qwen backbones, this compact interface outperforms the strongest evaluated fusion or token-compression baselines by about 2.9 correctness points on average.