AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

EMBGuard: Constructing Hazard-Aware Guardrails for Safe Planning in Embodied Agents

Dongwook Choi, Taeyoon Kwon, Bogyung Jeong, Minju Kim, Yeonjun Hwang, Hyojun Kim, Byungchul Kim, Young Kyun Jang, Jinyoung Yeo

ICML 2026 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Embodied agent learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
fine-tune Qwen-3-VL-2B on EMBHazard (EMBGuard-2B)Qwen-3-VL-2B · EMBGuardTestPotential Risk accuracy: 47.2 -> 51.6Improved
fine-tune Qwen-3-VL-4B on EMBHazard (EMBGuard-4B)Qwen-3-VL-4B · EMBGuardTestPotential Risk accuracy: 47.3 -> 54.3Improved
fine-tune Qwen-3-VL-4B on EMBHazard (EMBGuard-4B)Qwen-3-VL-4B · EMBGuardTestRisk Type accuracy: 51.0 -> 50.3No clear effect
fine-tune Qwen-3-VL-2B on EMBHazard (EMBGuard-2B)Qwen-3-VL-2B · Held-out SetRisk Type accuracy: 32.5 -> 59.5Improved
fine-tune Qwen-3-VL-4B on EMBHazard (EMBGuard-4B)Qwen-3-VL-4B · Held-out SetPotential Risk accuracy: 58.3 -> 71.2Improved
fine-tune InternVL3.5-1B on EMBHazardInternVL3.5-1B · EMBGuardTestPotential Risk accuracy gain: +37.6Improved
fine-tune Gemma-3-4B on EMBHazardGemma-3-4B · EMBGuardTestPotential Risk accuracy gain: +57.3Improved
scale guardrail from 2B to 4BEMBGuard · Held-out SetHazard accuracy: 36.6 -> 50.1Improved
scale guardrail from 2B to 4BEMBGuard · IS-Bench (pre-type process safety)F1: 30.5 -> 38.3Improved
provide correct risk type and hazard to policy modelpolicy models (GPT-4o, GPT-4o-mini, Qwen-3-VL-32B) · IS-Bench mitigation alignmentalignment with ground-truth strategy (%): 28.4 -> 90.4 (both wrong vs both correct); 78.5 wrong risk type; 58.6 wrong hazardImproved
use general MLLM as guardrail instead of specialized EMBGuard (false positive on benign scenarios)Gemini-2.5-Pro · EMBGuardTest benign scenariosfalse positive rate: 83.3% of benign scenarios misclassified as riskyHurt
freeze the vision encoder during fine-tuning (ablation)EMBGuard · preliminary experimentshazard explanation qualityImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy