# BLM-Guard: Explainable Multimodal Ad Moderation with Chain-of-Thought and Policy-Aligned Rewards

arXiv: https://arxiv.org/abs/2602.18193 · HTML: https://mlsanity.com/paper/2602.18193/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 24 of its 32 references that are in the map.
- First posted on arXiv 20 Feb 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)

## Builds on (most cited first, 24 of 24)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Tree of Thoughts: Deliberate Problem Solving with Large Language Models](https://mlsanity.com/paper/2305.10601/) (2023)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://mlsanity.com/paper/2312.06674/) (2023)
- [Kimi-VL Technical Report](https://mlsanity.com/paper/2504.07491/) (2025)
- [Adaptive Keyframe Sampling for Long Video Understanding](https://mlsanity.com/paper/2502.21271/) (2025)
- [Generative Language Models and Automated Influence Operations: Emerging Threats and Potential Mitigations](https://mlsanity.com/paper/2301.04246/) (2023)
- [GuardAgent: Safeguard LLM Agents by a Guard Agent via Knowledge-Enabled Reasoning](https://mlsanity.com/paper/2406.09187/) (2024)
- [GuardReasoner: Towards Reasoning-based LLM Safeguards](https://mlsanity.com/paper/2501.18492/) (2025)
- [MM-RLHF: The Next Step Forward in Multimodal LLM Alignment](https://mlsanity.com/paper/2502.10391/) (2025)
- [UnsafeBench: Benchmarking Image Safety Classifiers on Real-World and AI-Generated Images](https://mlsanity.com/paper/2405.03486/) (2024)
- [LlavaGuard: An Open VLM-based Framework for Safeguarding Vision Datasets and Models](https://mlsanity.com/paper/2406.05113/) (2024)
- [Interleaved-Modal Chain-of-Thought](https://mlsanity.com/paper/2411.19488/) (2024)
- [GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning](https://mlsanity.com/paper/2506.16141/) (2025)
- [On the Hardness of Faithful Chain-of-Thought Reasoning in Large Language Models](https://mlsanity.com/paper/2406.10625/) (2024)
- [GuardReasoner-VL: Safeguarding VLMs via Reinforced Reasoning](https://mlsanity.com/paper/2505.11049/) (2025)
- [FakeSV: A Multimodal Benchmark with Rich Social Context for Fake News Detection on Short Video Platforms](https://mlsanity.com/paper/2211.10973/) (2022)
- [SafeWatch: An Efficient Safety-Policy Following Video Guardrail Model with Transparent Explanations](https://mlsanity.com/paper/2412.06878/) (2024)
- [VLM as Policy: Common-Law Content Moderation Framework for Short Video Platform](https://mlsanity.com/paper/2504.14904/) (2025)
- [Moderating Illicit Online Image Promotion for Unsafe User-Generated Content Games Using Large Vision-Language Models](https://mlsanity.com/paper/2403.18957/) (2024)
- [T2Vs Meet VLMs: A Scalable Multimodal Dataset for Visual Harmfulness Recognition](https://mlsanity.com/paper/2409.19734/) (2024)
- [MSR-Align: Policy-Grounded Multimodal Alignment for Safety-Aware Reasoning in Vision-Language Models](https://mlsanity.com/paper/2506.19257/) (2025)

## Built on by (most cited first, 3 of 3)

- [RuleSafe-VL: Evaluating Rule-Conditioned Decision Reasoning in Vision-Language Content Moderation](https://mlsanity.com/paper/2605.07760/) (2026)
- [ARGUS: Policy-Adaptive Ad Governance via Evolving Reinforcement with Adversarial Umpiring](https://mlsanity.com/paper/2605.02200/) (2026)
- [SemTree: Judge-Aware Semantic Decision Trees for Inspectable and Repairable Rule Decisions](https://mlsanity.com/paper/2610.13999/) (2026)

## Abstract (arXiv)

Short-video platforms now host vast multimodal ads whose deceptive visuals, speech and subtitles demand finer-grained, policy-driven moderation than community safety filters. We present BLM-Guard, a content-audit framework for commercial ads that fuses Chain-of-Thought reasoning with rule-based policy principles and a critic-guided reward. A rule-driven ICoT data-synthesis pipeline jump-starts training by generating structured scene descriptions, reasoning chains and labels, cutting annotation costs. Reinforcement learning then refines the model using a composite reward balancing causal coherence with policy adherence. A multitask architecture models intra-modal manipulations (e.g., exaggerated imagery) and cross-modal mismatches (e.g., subtitle-speech drift), boosting robustness. Experiments on real short-video ads show BLM-Guard surpasses strong baselines in accuracy, consistency and generalization.