AI safety and alignment

Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.

17,884 stories · RSS feed

arXiv AI

Sep 1

Hyper3-CLIP: Hierarchy-Conditioned Hyperbolic Vision-Language Training

arXiv:2608.29313v1 Announce Type: cross Abstract: CLIP-like vision-language models (VLMs) trained with contrastive objectives learn strong global image-text representations, but their Euclidean embed...

By Matin Mahmood, Antonio Rueda-Toicen, Mohamed ElBassat, Seifeldin Elkerdany, Weixing Wang, Gerard de Melo

llms rag multimodal safety

arXiv AI

Sep 1

Scalable Clinical Data Infrastructure and Comparative ML Evaluation for Hospitalisation Risk Prediction in Elderly Patients with Multiple Long-Term Conditions using CPRD

arXiv:2608.29419v1 Announce Type: cross Abstract: Deep learning architectures are increasingly proposed for patient trajectory modeling in electronic health records (EHRs), yet their advantage over s...

By Asra Aslam, Volodymyr Chapman, Maurice M. O'Connell, Aseel S. Abuzour, Michael Abaho, Danushka Bollegala, Gary Leeming, Eduard Shantsila, Andrew Clegg, Lauren E. Walker, Iain Edward Buchan, Samuel D. Relton

benchmarks safety

arXiv AI

Sep 1

Applications of Risk Science to AI Fairness Evaluation: Principles, Challenges, and Best Practices

arXiv:2608.29478v1 Announce Type: cross Abstract: Scholarly work which aims to describe potential societal impacts (e.g., risks) of proliferating technology (especially related to artificial intellig...

By Kyra Wilson, Sabrina Kang, Saloni Dash, Aylin Caliskan

safety

arXiv AI

Sep 1

Argument-Aware Semantic Alignment of Normative Texts: A Toulmin-Based Neuro-Symbolic Approach

arXiv:2608.29529v1 Announce Type: cross Abstract: Semantic alignment between specialized normative texts is challenging when equivalent requirements use different terms, syntax, and levels of abstrac...

By William Schroeder

llms rag benchmarks safety

arXiv AI

Sep 1

Memory-First Fact-Checking: A Knowledge-Graph-Grounded Multi-Agent System for Misinformation Detection

arXiv:2608.29617v1 Announce Type: cross Abstract: This paper introduces a hybrid fact-checking framework that integrates Knowledge Graph-based semantic memory with adversarial multi-agent reasoning f...

By Amelia Petrenciuc, Alexandru Lecu, Adrian Groza

llms agents benchmarks safety

arXiv AI

Sep 1

R$^2$A: Learning Persona Policies Through Persona Representation Learning and Runtime Alignment

arXiv:2608.29798v1 Announce Type: cross Abstract: The same Persona behavior can be beneficial in one context but harmful in another, causing static Persona elicitation to perform inconsistently acros...

By Mohan Zhang, Chengsong You, Xiaoyu Cao, Zhen Sun, Xiaohan Jia, Junwei Zhou, Yongchao Chen

benchmarks safety

arXiv AI

Sep 1

Influence Is Not Authority: When Causal Guardrail Signals Make Legitimate Tool Use Look Like an Attack in Tool-Using LLM Agents

arXiv:2608.29942v1 Announce Type: cross Abstract: The key limitation of current state-of-the-art influence-based guardrails is that they do not reliably distinguish a legitimate, user-authorized acti...

By Tanzim Ahad, Ismail Hossain, Md Jahangir Alam, Sai Puppala, Syed Bahauddin Alam, Sajedul Talukder

llms agents benchmarks safety

arXiv AI

Sep 1

Partition-Aware Unlearning for Removing Spurious Correlations in Large Vision-Language Models

arXiv:2608.29996v1 Announce Type: cross Abstract: Large Vision-Language Models (LVLMs) achieve strong performance across many multimodal tasks; however, they often exploit spurious object-background...

By Aditi Sarker, Nazreen Shah, Rafi Ibn Sultan, Rhongho Jang, Dongxiao Zhu, Prashant Khanduri

llms multimodal benchmarks safety

arXiv AI

Sep 1

"Act Like a 5th Grader" is Not Enough: Bounding Knowledge in LLM-Based User Simulators

arXiv:2608.30033v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used to simulate human behavior but frequently fail to exhibit realistic cognitive constraints, sufferi...

By Krisztian Balog, Arild Michel Bakken

llms safety

arXiv AI

Sep 1

Pak3H: Evaluating the Cost of Cultural Mismatch in LLM Alignment with a Human-Contextualized Urdu Benchmark

arXiv:2608.30065v1 Announce Type: cross Abstract: Large language models (LLMs) demonstrate strong Helpfulness, Harmlessness, and Honesty (3H) alignment in English-centric settings, but these gains tr...

By Abdullah Hashmat, Usman Naseem, Agha Ali Raza

llms benchmarks safety

arXiv AI

Sep 1

VIBE: Video Instruction-aligned Background music gEneration

arXiv:2608.30125v1 Announce Type: cross Abstract: Current video-to-music (V2M) models lack semantic control and fail to penalize instruction violations, largely due to their reliance on reconstructio...

By Aryan Vijay Bhosale, Vaibhavi Lokegaonkar, Vishnu Raj, Gouthaman KV, Sreyan Ghosh, Ramani Duraiswami, Lie Lu, Dinesh Manocha

diffusion reinforcement-learning multimodal safety

arXiv AI

Sep 1

SIR: Self-improving Red-teaming for Compute Use Agents

arXiv:2608.30207v1 Announce Type: cross Abstract: Computer use agents (CUAs) are vision-language models that perceive a screen and act on a real operating system through mouse, keyboard, and terminal...

By Chen Xiong, Zhiyuan He, Pin-Yu Chen, Stjepan Picek, Tsung-Yi Ho

llms agents multimodal benchmarks safety

arXiv AI

Sep 1

Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLMs via Cross-Family Representation Steering

arXiv:2608.30319v1 Announce Type: cross Abstract: Large language models (LLMs) finetuned for specialized domains represent crucial high-impact applications. Inference-time alignment improves safety d...

By Jin Gan, Xin Li, Jun Luo

llms fine-tuning benchmarks safety

arXiv AI

Sep 1

SemPOI-RL: Aligning LLM Semantic Reasoning for Interpretable Out-of-Town POI Sequential Generation

arXiv:2608.30399v1 Announce Type: cross Abstract: Large language models (LLMs) exhibit strong semantic reasoning and open-ended generation abilities, but aligning these abilities with structured sequ...

By Yunqi Liu, Yang Zhang, Ruixing Zhang, Liangzhe Han, Yi Qiao, Tongyu Zhu, Leilei Sun

llms reinforcement-learning fine-tuning safety

arXiv AI

Sep 1

Trajectory-Initialized Neural Double Q-Routing for Large-Scale Overhead Hoist Transport Systems

arXiv:2608.30512v1 Announce Type: cross Abstract: Large-scale industrial robot fleets share constrained physical infrastructure, making vehicle travel times dependent on safety separation, intersecti...

By Cheng Gu, Qiusheng Zhao, Anbang Liu, Shaochong Lin, Max Z. J. Shen

robotics safety

arXiv AI

Sep 1

Collapsibility of Performance Metrics in Clinical Predictive AI

arXiv:2608.30568v1 Announce Type: cross Abstract: Background: Population level assessments of predictive artificial intelligence (AI) can conceal performance disparities across subgroups. Fairness ev...

By Jo\~ao Matos, Ben Van Calster, Richard D. Riley, Paula Dhiman, Gary S. Collins

safety

arXiv AI

Sep 1

Learning Dynamics of Logits Debiasing for Long-Tailed Semi-Supervised Learning

arXiv:2608.30699v1 Announce Type: cross Abstract: Long-tailed distributions are prevalent in real-world semi-supervised learning (SSL), where pseudo-labels tend to favor majority classes, leading to...

By Yue Cheng, Jiajun Zhang, Xiaohui Gao, Weiwei Xing, Zhanxing Zhu

efficiency benchmarks safety

arXiv AI

Sep 1

RailGen: Improving Railway Intrusion Detection via Agent-Guided Small-Scale Foreign Object Generation

arXiv:2608.30727v1 Announce Type: cross Abstract: Small-object detection under long-tailed data distributions is a fundamental yet challenging problem in multimedia. Railway Foreign Object Detection...

By Quan Hao, Ziyang Tao, Chenxi Zhang, Yudong Wang, Rui Shi, Liguo Zhang

agents diffusion computer-vision multimodal benchmarks safety

arXiv AI

Sep 1

Lucida: Parse, Generate, and Place for Composable Real-to-Sim Scene Modeling

arXiv:2608.30821v1 Announce Type: cross Abstract: Composable scene modeling aims to recover a real indoor scene as complete, editable object assets arranged as observed, giving robot simulation and e...

By Minghan Qin, Yuang Wang, Xiuyu Yang, Yushi Long, Yujian Zhang, Ruihuan Wang, Kai Ye, Yangang Zhang, Hang Li

computer-vision robotics multimodal safety

arXiv AI

Sep 1

TAMI: Temporally Aligned, Missingness-Aware, and Interpretable Multimodal Fusion for Mental Health Assessment in Older Adults with Mild Cognitive Impairment

arXiv:2608.30857v1 Announce Type: cross Abstract: Depression and anxiety in older adults with Mild Cognitive Impairment (MCI) are frequently underdiagnosed due to limited access to care. Multimodal a...

By Merna Bibars, Bolaji Omofojoye, Allan I. Levey, Rachel Hershenberg, Gari D. Clifford, Hyeokhyen Kwon

multimodal safety

Get the daily briefing