Safety Benchmarks

Follow this topic Ask about this topic

Momentum

7 papers in the last four weeks, against 2 the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 82

All topics

Multimodal Safety Evaluation Should Measure Controllability Beyond Classification
Oct 5, 2026Junhyeong Park, Hanwool Lee, DongGeon Lee +4Safety Benchmarks Multimodal Evaluation
KAIST · AIM Intelligence · Seoul National University +3
SCOUT: Synergizing Reasoning and Tool-Use for Computer-Use Safety
Sep 28, 2026Jianxing Chen, Xiao Yu, Shipra Agrawal +1Safety Benchmarks Security Evaluation
Columbia University
AnesTRACE: Benchmarking Intraoperative Anesthesia from Multimodal Perception to Multi-step Decision-Making
Sep 26, 2026Ziwei Huang, Qi Gao, Zhe Ji +5Safety Benchmarks Perception
School of Software Technology, Zhejiang University · Department of Anesthesiology, The Second Affiliated Hospital of Zhejiang University School of Medicine · School of Software, Central South University +1
CasCVS-Net: A Staged Multi-Task Cascade for Critical View of Safety Assessment
Sep 23, 2026Bock-Zien Toh, Yuanchuan Ren, Tay Aw Yu +3Colonoscopy Computer Vision
Department of Computer Science, University College London, London, UK · UCL Hawkes Institute, University College London, UK
Evaluation Metrics for Safe Reinforcement Learning
Sep 14, 2026Lindsay Spoor, Aske Plaat, Thomas MoerlandSafety Benchmarks Markov Decision Processes
Leiden Institute of Advanced Computer Science, Leiden University, Leiden, the Netherlands
Scenario-Independent Criticality Assessment and Prediction for Vulnerable Road Users in Autonomous Driving
Sep 14, 2026Jörg Gamerdinger, Victor Schwarzenberger, Philipp Schmid +2Automated Vehicles Autonomous Driving
University of Tübingen, Faculty of Science, Department of Computer Science, Embedded Systems Group
BlueprintAgent: Constraint-Triggered Targeted Revisits for Simulation-Ready Generation from Scanned Structural Blueprints
Sep 7, 2026Zhouyuan Xu, Chen Yang, Linhao Wang +2Building Information Modeling Engineering
Tsinghua University, Beijing 100084, China
How Long Until Your Robot Ignores You? A Safety Benchmark for LLM Orchestrators in Human-Humanoid Collaboration
Sep 7, 2026Aulon Bajrami, Mohamed Elshamouty, Werner KrausSafety Benchmarks Large Language Model Workflows
Fraunhofer IPA Stuttgart, Germany
ObserverBench: Testing Mechanistic Estimates for Intervention and Control
Sep 2, 2026Vijay ErramilliMechanistic Interpretability Safety Benchmarks
Beyond Semantic Accuracy: Consequence-Aware Evaluation for Safety-Critical Language Understanding
Aug 25, 2026Yujing Chang, Thinh Pham, Van-Phat Thai +4Safety Benchmarks Large Language Model Safety
1ATMRI, Nanyang Technological University (NTU), Singapore · Centre of AI Research, VinUniversity, Vietnam · School of Management, Indian Institute of Technology Mandi, India
SurakshaEval: An Indic Safety Benchmark for Multilingual LLMs
Aug 8, 2026Debopriyo Banerjee, Kapil Rajesh Kavitha, Angana Borah +11Multilingual Safety Large Language Model Safety
Inception · Mohamed bin Zayed University of Artificial Intelligence · University of Michigan +4
How Should I Pick a Foundation Model for My Robot? In Favor of a Community Evaluation Framework for Social Robots
Aug 7, 2026Eric Nichols, Alva Markelius, Hatice GunesMulti-Dimensional Evaluation Safety Benchmarks
Honda Research Institute Japan, Wako, Japan. · Affective Intelligence and Robotics Lab, University of Cambridge, UK.
Item Response Theory for AI Safety
Aug 5, 2026Joshua Fonseca Rivera, Neil Shah, David Demitri Africa +1Item Response Theory Safety Benchmarks
Independent · 1Independent · UK AI Security Institute +1
ADMITBench: A Safety-Governed Reference Framework for Evaluating the Admissibility of Industrial LLM Advisories
Aug 4, 2026Yash Misra, Javal Vyas, Siddharth Gutta +1Safety Benchmarks Safety Claims
Refiant, Inc. · Imperial College London
Evaluating Counterfactual Sensitivity to Patient Information in Medication-Safety Reasoning
Aug 4, 2026Zhitian Hou, Yuhang Liu, Pengkai Wang +8Medication Leaflet Patients
The Hong Kong Polytechnic University · Sun Yat-sen University
When May a Model Replace the Experiment? Audits, Licenses, and the Price of Trust in Surrogate-Driven Design
Aug 2, 2026Shuangxiu, Ma, Wenhe +1Model Auditing Surrogate Gradient
Max · William G. Lowrie Department of Chemical and Biomolecular Engineering, The Ohio State University, Columbus, Ohio 43210, United States · Zachary +1
OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution
Aug 1, 2026Yunhao Chen, Xin Wang, Yixu Wang +6Red-Teaming Artificial Intelligence Safety
Fudan University · Shanghai Artiﬁcial Intelligence Laboratory · Work done during Yunhao Chen’s internship at Shanghai AI Lab. +1
InfoOps Bench: A live information operations safety benchmark
Jul 30, 2026Dorian Quelle, Lisa-Maria Neudert, Jonathan Bright +1Artificial Intelligence Benchmarks Fact-Checking
Safety, or Just Capability? A Validity Audit of Agent-Safety Benchmarks
Jul 30, 2026Youting Wang, Xiao Han, Dingyan Shang +2Safety Benchmarks Security Evaluation
SafeBuild-Bench: A Temporal-Robust Construction Safety Benchmark with Graph-Enhanced Data Mining
Jul 29, 2026Yi Cui, Zilin Wang, Yijie Xu +5Safety Benchmarks Hazard
Hong Kong University of Science and Technology (Guangzhou) Guangzhou, China · The Hong Kong University of Science and Technology(Guangzhou) Guangzhou, China · School of Management Northwest Polytechnical University Xi’an Xi’an, China +4
What AI Red-Team Evaluations Can and Cannot Prove
Jul 23, 2026Bandana KaurArtificial Intelligence Evaluation Safety Benchmarks
APIsec Research Labs
QuantiBias: Benchmarking Quantization-Induced Bias in LLMs
Jul 23, 2026Emilio FerraraLarge Language Model Bias Large Language Model Quantization
Thomas Lord Department of Computer Science University of Southern California
OpenSkillRisk: Benchmarking Agent Safety When Using Real-World Risky Third-Party Skills
Jul 22, 2026Qiyuan Liu, Tingfeng Hui, Kun Zhan +2Reusable Agent Skills Skills
Department of Data Science, City University of Hong Kong · Hong Kong Institute of AI for Science, City University of Hong Kong · Beijing University of Posts and Telecommunications +1
Clinical Pathways as Safety Specifications for Physical AI in Hospital Wards
Jul 22, 2026Gabriele Franchini, Giulio Mallardi, Michele De Carolis +1Healthcare Safety Benchmarks
Computer Science Dept. University of Bari Bari, Italy
SciHazard: A Benchmark for Measuring Scientific Safety Risks with Decomposed Harm Scoring
Jul 21, 2026Chunxiao Li, Yuan Xiong, Lijun Li +4Safety Benchmarks Hazard
Shanghai Artificial Intelligence Laboratory
SinD 2.0: A Multi-City UAV Dataset with Semantic Risk Annotations for SOTIF-Oriented Safety Validation at Signalized Intersections
Jul 18, 2026Yunwei Li, Shengjie Fu, Chunrong Chen +16Naturalistic Driving Data Aerial Dataset
SafeRelBench: A Spatial-Relation-Aware Benchmark for Process-Level Safety in VLM-Driven Embodied Agents
Jul 16, 2026Huaigang Yang, Ya Li, Min Ren +3Safety Benchmarks Embodied Agents
Beijing University of Posts and Telecommunications, Beijing, China · State Key Laboratory of General Artificial Intelligence, BIGAI, Beijing, China
A Threshold Exceedance Framework for CBRN Uplift Evaluation in Frontier Language Models
Jul 13, 2026Rahul Gupta, Abhinav Mohanty, Payal Motwani +8Realistic Threat Model Rise
Amazon Nova Responsible AI · Work done while at Amazon. · Nemesys Insights
Safety Testing LLM Agents at Scale: From Risk Discovery to Evidence-Grounded Verification
Jul 2, 2026Yunhao Feng, Ruixiao Lin, Ming Wen +12Safety Benchmarks
AntGroup · Hunan Institute of Advanced Technology · Zhejiang University +3
EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety Guards
Jun 30, 2026Siddhant Panpatil, Arth Singh, Mijin Koo +3Safety Benchmarks Egocentric Video
AIM Intelligence · Seoul National University
OopsieVerse: A Safety Benchmark with Damage-Aware Simulation for Robot Manipulation
Jun 30, 2026Arnav Balaji, Arpit Bahety, Sriniket Ambatipudi +3Robotic Manipulation Safety Benchmarks
The University of Texas at Austin
Uncertainty-Aware Last-Layer Adaptation of RETFound for Referable Diabetic Retinopathy Screening Under Dataset Shift
Jun 30, 2026Karim MardhaniGlaucoma Detection Uncertainty-Aware Classification
University of Colorado Boulder Boulder, CO, USA
EMPATH: A Multilingual Auditor-Judge Benchmark for Safety Evaluation of Emotional-Support Chatbots
Jun 29, 2026Camilo Chacón SartoriEmotional Support Conversation Task Empathy
MindSurf
EvalSafetyGap: A Hybrid Survey and Conceptual Framework for LLM Evaluation-Safety Failures
Jun 29, 2026Buğra Alperen Uluırmak, Rifat KurbanSafety Benchmarks Model Auditing
Erciyes University, Türkiye. · Abdullah Gül University, Türkiye.
SafePyramid: A Hierarchical Benchmark for In-context Policy Guardrailing
Jun 29, 2026Jiacheng Zhang, Haoyu He, Sen Zhang +5Safety Benchmarks
1ByteDance · 2The University of Melbourne
Can LLMs Reliably Self-Report Adversarial Prefills, and How?
Jun 22, 2026Quang Minh Nguyen, Uzair Ahmed, Taegyoon KimLarge Language Model Safety Introspection Adapters
KAIST
Evaluation Awareness Is Not One Capability: Evidence from Open Language Models
Jun 22, 2026Nilesh Nayan, Aishwarya Sampath Kumar, Rishiraj Girmal +5Safety Benchmarks
University of Massachusetts, Amherst · 2Microsoft Corp.
NRT-Bench: Benchmarking Multi-Turn Red-Teaming of LLM Operator Agents in Safety-Critical Control Rooms
Jun 18, 2026Hanwool Lee, Dasol Choi, Bokyeong Kim +2Red-Teaming Safety Benchmarks
AIM Intelligence · Republic of Korea · KAERI
OSGuard: A Benchmark for Safety in Computer-Use Agents
Jun 13, 2026Mina Mohammadmirzaei, Jeffrey FlaniganSafety Benchmarks Computer-Use Agents
University of California, Santa Cruz
Quality Is Not a Safety Proxy Under Quantization
Jun 8, 2026Sahil KadadekarQuantized Safety Benchmarks
Independent Researcher
Reliable to Expressive: A Curriculum for Rubric-Following Safety Judges
Jun 8, 2026Yongtaek Lim, Hyeji Choi, Minwoo KimSafety Benchmarks Judges
AI Safety Team, DATUMO.INC, Seoul, South Korea.
The Cold-Start Safety Gap in LLM Agents
Jun 5, 2026Chung-En Sun, Linbo Liu, Tsui-Wei WengLarge Language Model Agents Large Language Model Safety
University of California, San Diego
DOG-DPO:Dynamic Optimization in Geometry for Safety Alignment
Jun 4, 2026Yi Nian, Tiankai Yang, Yudi Zhang +7Safety Alignment Large Language Model Alignment
University of Southern California · Iowa State University · University of Wisconsin–Madison +4
Attack Selection in Agentic AI Control Evaluations Meaningfully Decreases Safety
Jun 3, 2026Catherine Ge-Wang, Tyler Crosse, Benjamin Hadad +3Artificial Intelligence Safety Artificial Intelligence Control
University of Oxford · Mentorship for Alignment Research Students (MARS), CAISH · Georgia Institute of Technology +3
What Benchmarks Don't Measure: The Case for Evaluating Abstention Competence in Autonomous Agents
Jun 1, 2026Victor Ojewale, Suresh VenkatasubramanianAgentic Evaluations Autonomous Agents
Brown University Providence, Rhode Island, USA
LLM Judges Inconsistently Disagree Across Safety Criteria and Harm Categories
May 29, 2026Krishnapriya Vishnubhotla, Sowmya Vajjala, Akriti Vij +1Large Language Model Judges Judges
National Research Council, Canada · IMDA, Singapore
Configurable Reward Model for Balanced Safety Alignment
May 28, 2026Zhengping Jiang, Mehran Khodabandeh, Akash Bharadwaj +5Safety Alignment Large Language Model Alignment
Johns Hopkins University · Meta · Meta Superintelligence Labs
When Context Flips, Safety Breaks: Diagnosing Brittle Safety in Aligned Language Models
May 27, 2026Dasol Choi, Alex KwonLarge Language Model Safety Safety Benchmarks
AIM Intelligence · Independent Researcher
Got a Secret? LLM Agents Can't Keep It: Evaluating Privacy in Multi-Agent Systems
May 26, 2026Aman Priyanshu, Supriti Vijay, Esha PahwaLarge Language Model Agents Artificial Intelligence Agents
Foundation AI · USA · Corvic AI
Efficient Safety Benchmarking via Item Response Theory
May 26, 2026Fabio Spagliardi, Mírian Silva, Ayan Datta +3Safety Benchmarks Item Response Theory
Supervised Program for Alignment Research (SPAR) · Yale University · Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)
Semantic Robustness Probing via Inpainting: An Interactive Tool for Safety-Critical Object Detection
May 26, 2026Nico Steckhan, Krutarth Prajapati, Weija Shao +1Inspection Inpainting
Federal Institute for Occupational Safety and Health (BAuA), Germany
Decomposing and Measuring Evaluation Awareness
May 21, 2026Changling Li, Terry Jingchen Zhang, Jie Zhang +4Safety Benchmarks Frontier Models
ETH Zürich · Max Planck Institute for Intelligent Systems · University of Toronto & Vector Institute +3
Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety
May 21, 2026Piercosma Bisconti, Matteo Prandi, Federico Pierucci +11Large Language Model Safety Safety Benchmarks
Icaro Foundation · Sapienza University of Rome · Sant’Anna School of Advanced Studies +6
Guiding Neuro-Symbolic Scenario Generation with Spatio-Temporal Logic
May 18, 2026Lorenzo Bonin, Francesco Giacomarra, Luca Bortolussi +2Scenario Generation Safety-Critical Scenarios
University of Trieste · University of Southern California
Reinterpreting Safety Thresholds as Neuron Spiking Thresholds
May 18, 2026Enrico Del Re, Mohamed Sabry, Cristina Olaverri-MonrealEmergency Braking Intensity Prediction Spiking Neural Networks
Johannes Kepler University Linz, Austria; Department Intelligent Transport Systems
REBAR: Reference Ethical Benchmark for Autonomy Readiness
May 18, 2026Jonathan Diller, David Barnes, Rebekah Bogdanoff +14Ethics Safety Benchmarks
University of Pennsylvania · David Barnes, LLC · Duality Robotics, Inc. +3
DisaBench: A Participatory Evaluation Framework for Disability Harms in Language Models
May 12, 2026Eugenia Kim, Ioana Tanase, Christina MallonSafety Benchmarks Disability
Microsoft
Safety-Oriented Evaluation of Language Understanding Systems for Air Traffic Control
May 12, 2026Yujing Chang, Yash Guleria, Duc-Thinh Pham +4Air Traffic Control Safety Benchmarks
1ATMRI, Nanyang Technological University (NTU), Singapore · School of Management, Indian Institute of Technology Mandi, India · Centre of AI Research, VinUniversity, Vietnam
Navigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks
May 11, 2026Regina Gugg, Selina Niederländer, Andreas Stöckl +1Large Language Model Evaluation Toxicity
Dynatrace Research, Linz, Austria · University of Applied Sciences Upper Austria, Hagenberg, Austria
IndustryBench: Probing the Industrial Knowledge Boundaries of LLMs
May 11, 2026Songlin Bai, Xintong Wang, Linlin Yu +12Safety Benchmarks Large Language Model Evaluation
Multimodal and Industrial AI Team† Taobao&Tmall, Alibaba Group

Oct 5, 2026, cs.AIJ/K move · Enter open · S save

Multimodal Safety Evaluation Should Measure Controllability Beyond Classification

Junhyeong Park, Hanwool Lee, DongGeon Lee, Dasol Choi+3

KAIST · AIM Intelligence · Seoul National University+3

Safety Benchmarks Multimodal Evaluation

Open paper AskPDF arXiv

VLM safety is commonly evaluated through input- and output-level classification. Such classification is necessary, but it does not reveal whether a safety state is accessible or controllable inside the model. We argue that multimodal safety evaluation should therefore report a \emph{controllability profile} alongside behavioral classification, separating representation-level detectability, cross-modal specificity, intervention sensitivity, and benign-preserving selectivity. Using implicit toxicity as a stress case, we instantiate this profile on LlavaGuard and Qwen3.5 with sparse feature decompositions. LlavaGuard admits localized handles with a narrow benign-preserving intervention range and modest downstream safety gains, whereas Qwen3.5 supports strong representation-level readout but no comparable selective-control regime under the tested operators. These results show that internal readout and controllability can diverge. Future multimodal safety benchmarks should therefore report not only behavioral safety metrics, but also whether safety-relevant internal signals can be intervention-tested and controlled within a validated operating range.

Momentum

Related topics

Latest papers 82