# Fake-HR1: Rethinking Reasoning of Vision Language Model for Synthetic Image Detection

arXiv: https://arxiv.org/abs/2602.10042 · HTML: https://mlsanity.com/paper/2602.10042/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 29 of its 33 references that are in the map.
- First posted on arXiv 10 Feb 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [MLLM jailbreaks and adversarial robustness](https://mlsanity.com/mllm/mllm-jailbreaks-and-adversarial-robustness/)

## Builds on (most cited first, 29 of 29)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [Qwen-Image Technical Report](https://mlsanity.com/paper/2508.02324/) (2025)
- [WorldSimBench: Towards Video Generation Models as World Simulators](https://mlsanity.com/paper/2410.18072/) (2024)
- [Statistical Rejection Sampling Improves Preference Optimization](https://mlsanity.com/paper/2309.06657/) (2023)
- [GenImage: A Million-Scale Benchmark for Detecting AI-Generated Image](https://mlsanity.com/paper/2306.08571/) (2023)
- [RADAR: Robust AI-Text Detection via Adversarial Learning](https://mlsanity.com/paper/2307.03838/) (2023)
- [A Sanity Check for AI-generated Image Detection](https://mlsanity.com/paper/2406.19435/) (2024)
- [FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models](https://mlsanity.com/paper/2410.02761/) (2024)
- [Thinkless: LLM Learns When to Think](https://mlsanity.com/paper/2505.13379/) (2025)
- [Spot the Fake: Large Multimodal Model-Based Synthetic Image Detection with Artifact Explanation](https://mlsanity.com/paper/2503.14905/) (2025)
- [Think Only When You Need with Large Hybrid-Reasoning Models](https://mlsanity.com/paper/2505.14631/) (2025)
- [Common Sense Reasoning for Deepfake Detection](https://mlsanity.com/paper/2402.00126/) (2024)
- [LOKI: A Comprehensive Synthetic Data Detection Benchmark using Large Multimodal Models](https://mlsanity.com/paper/2410.09732/) (2024)
- [Community Forensics: Using Thousands of Generators to Train Fake Image Detectors](https://mlsanity.com/paper/2411.04125/) (2024)
- [Towards General Visual-Linguistic Face Forgery Detection](https://mlsanity.com/paper/2307.16545/) (2023)
- [IMDL-BenCo: A Comprehensive Benchmark and Codebase for Image Manipulation Detection & Localization](https://mlsanity.com/paper/2406.10580/) (2024)
- [FakeBench: Probing Explainable Fake Image Detection via Large Multimodal Models](https://mlsanity.com/paper/2404.13306/) (2024)
- [OmniGuard: Hybrid Manipulation Localization via Augmented Versatile Deep Image Watermarking](https://mlsanity.com/paper/2412.01615/) (2024)
- [Mesoscopic Insights: Orchestrating Multi-scale & Hybrid Architecture for Image Manipulation Localization](https://mlsanity.com/paper/2412.13753/) (2024)
- [Ivy-Fake: A Unified Explainable Framework and Benchmark for Image and Video AIGC Detection](https://mlsanity.com/paper/2506.00979/) (2025)
- [DiffusionFake: Enhancing Generalization in Deepfake Detection via Guided Stable Diffusion](https://mlsanity.com/paper/2410.04372/) (2024)
- [GLFF: Global and Local Feature Fusion for AI-synthesized Image Detection](https://mlsanity.com/paper/2211.08615/) (2022)
- [ForensicHub: A Unified Benchmark & Codebase for All-Domain Fake Image Detection and Localization](https://mlsanity.com/paper/2505.11003/) (2025)
- [UniShield: An Adaptive Multi-Agent Framework for Unified Forgery Image Detection and Localization](https://mlsanity.com/paper/2510.03161/) (2025)
- [SIDA: Synthetic Image Driven Zero-shot Domain Adaptation](https://mlsanity.com/paper/2507.18632/) (2025)

## Built on by (most cited first, 8 of 8)

- [Venus-DeFakerOne: Unified Fake Image Detection & Localization](https://mlsanity.com/paper/2605.14091/) (2026)
- [HyTRec: A Hybrid Temporal-Aware Attention Architecture for Long Behavior Sequential Recommendation](https://mlsanity.com/paper/2602.18283/) (2026)
- [Veritas++: Value-aware On-Policy Distillation for Perception-Enhanced AIGI Detection](https://mlsanity.com/paper/2607.27113/) (2026)
- [UniMoMo: Expert Merging-Based MoE Acceleration for Large Recommendation Models](https://mlsanity.com/paper/2608.08627/) (2026)
- [EditSleuth: A Dataset of Grounded Reasoning Chains for Image-Edit Forensics](https://mlsanity.com/paper/2605.08695/) (2026)
- [Can Vision-Language Models Reason about AI Edits in Images?](https://mlsanity.com/paper/2607.28464/) (2026)
- [UniAR: A Unified Framework for Autism Recognition Enhanced by Multi-View Prompt Learning](https://mlsanity.com/paper/2609.31298/) (2026)
- [PolyOCR-Venus: Unified OCR Foundation Models for Text-Centric Visual Intelligence](https://mlsanity.com/paper/2609.37712/) (2026)

## Abstract (arXiv)

Recent studies have demonstrated that incorporating Chain-of-Thought (CoT) reasoning into the detection process can enhance a model's ability to detect synthetic images. However, excessively lengthy reasoning incurs substantial resource overhead, including token consumption and latency, which is particularly redundant when handling obviously generated forgeries. To address this issue, we propose Fake-HR1, a large-scale hybrid-reasoning model that, to the best of our knowledge, is the first to adaptively determine whether reasoning is necessary based on the characteristics of the generative detection task. To achieve this, we design a two-stage training framework: we first perform Hybrid Fine-Tuning (HFT) for cold-start initialization, followed by online reinforcement learning with Hybrid-Reasoning Grouped Policy Optimization (HGRPO) to implicitly learn when to select an appropriate reasoning mode. Experimental results show that Fake-HR1 adaptively performs reasoning across different types of queries, surpassing existing LLMs in both reasoning ability and generative detection performance, while significantly improving response efficiency.