AI safety and alignment

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

10,811 stories · RSS feed

arXiv AI

Jun 8

TEVI: Text-Conditioned Editing of Visual Representations via Sparse Autoencoders for Improved Vision-Language Alignment

arXiv:2606. 07451v1 Announce Type: cross Abstract: Vision-language models such as CLIP are highly useful for diverse tasks due to their shared image-text embedding space.

By Sweta Mahajan, Sukrut Rao, Jiahao Xie, Alexander Koller, Bernt Schiele

llms rag multimodal benchmarks safety

arXiv AI

Jun 8

Re-imagining ISO 26262 in the Age of Autonomous Vehicles: Enhancing Controllability through Transferability and Predictability

arXiv:2606. 07437v1 Announce Type: cross Abstract: The ISO 26262 standard defines functional safety for road vehicles through risk assessments based on Severity, Exposure, and Controllability, grounded in a human-driven vehicle paradigm.

By Chaitanya Shinde, Hadi Hajieghrary, Paul Schmitt, Adam Shoemaker, Bodo Seifert, Steve Kenner

agents robotics safety

arXiv Machine Learning

Jun 8

Accelerating Reproducible Research in Synthetic EHR Generation

arXiv:2606. 06990v1 Announce Type: new Abstract: The generation of high-fidelity synthetic Electronic Health Records (EHR) is crucial for advancing medical research while preserving patient privacy.

By Jalen Jiang, Chufan Gao, Ethan Rasmussen, Stephen Z. Xie, Jimeng Sun

llms diffusion benchmarks safety

arXiv AI

Jun 8

PaperFlow: Profiling, Recommending, and Adapting Across Daily Paper Streams

arXiv:2606. 07454v1 Announce Type: cross Abstract: Scientific paper recommendation is typically evaluated as static ranking over a fixed candidate set, yet real scientific reading unfolds as a daily, longitudinal process in which interests shift and feedback accumulates.

By Fuqiang Wang, Song Tan, Zheng Guo, Jiaohao Fu, Xinglong Xu, Bihui Yu, Jie Dong, Zheng Sun, Siyuan Li, Jingxuan Wei, Cheng Tan

benchmarks safety

arXiv Machine Learning

Jun 8

SHAP-Guided Kernel Actor-Critic for Explainable Reinforcement Learning

arXiv:2512. 05291v3 Announce Type: replace Abstract: Actor-critic (AC) methods are a cornerstone of reinforcement learning (RL) but offer limited interpretability.

By Na Li, Hangguan Shan, Wei Ni, Wenjie Zhang, Xinyu Li

rag reinforcement-learning safety

arXiv Machine Learning

Jun 8

Attention Sink in Transformers: A Survey on Utilization, Interpretation, and Mitigation

arXiv:2604. 10098v2 Announce Type: replace Abstract: As the foundational architecture of modern machine learning, Transformers have driven remarkable progress across diverse AI domains.

By Zunhai Su, Hengyuan Zhang, Wei Wu, Yifan Zhang, Yaxiu Liu, He Xiao, Qingyao Yang, Yuxuan Sun, Rui Yang, Chao Zhang, Jing Xiong, Hui Shen, Keyu Fan, Weihao Ye, Chaofan Tao, Taiqiang Wu, Zhongwei Wan, Tiantian Zhang, Bowen Yan, Zhen Li, Yiming Zhang, Congkai Xie, Yulei Qian, Yuchen Xie, Yik-Chung Wu, Hongxia Yang, Ngai Wong

llms safety

arXiv AI

Jun 8

LoRA-DA: Data-Aware Initialization for Low-Rank Adaptation via Asymptotic Analysis

arXiv:2510. 24561v3 Announce Type: replace-cross Abstract: LoRA has become a widely adopted method for PEFT, and its initialization methods have attracted increasing attention.

By Qingyue Zhang, Chang Chu, Tianren Peng, Qi Li, Xiangyang Luo, Zhihao Jiang, Shao-Lun Huang

fine-tuning benchmarks safety

arXiv Machine Learning

Jun 8

Sparsely gated tiny linear experts

arXiv:2606. 07414v1 Announce Type: new Abstract: Sparsity allows scaling model parameters without proportionally increasing computational cost.

By Simon Schug

llms efficiency safety

arXiv AI

Jun 8

PandaAI: A Practical Agent CQ2 for Neuro-symbolic Data Analysis And Integrated Decision-Making in Quantitative Finance

arXiv:2606. 06823v1 Announce Type: cross Abstract: While deep learning has excelled in various domains, its application to sequential decision-making in finance remains challenging due to the low Signal-to-Noise Ratio (SNR) and non-stationarity of financial data.

By Yuqi Li, Siyuan Liu, Bingjun Liu

llms agents fine-tuning benchmarks safety

arXiv Machine Learning

Jun 8

Unmixing ATR-{\mu}FTIR spectroscopic images of cross-sections of historical oil paintings

arXiv:2603. 06673v2 Announce Type: replace-cross Abstract: Spectroscopic imaging (SI) has become central to heritage science because it enables non-invasive, spatially resolved characterisation of materials in artefacts.

By Shivam Pande, Nicolas Nadisic, Francisco Mederos-Henry, Aleksandra Pizurica

safety

arXiv Machine Learning

Jun 8

Explaining Unsupervised Disease Staging in Huntington's Disease: Insights into Model Representations and Clusters

arXiv:2606. 07135v1 Announce Type: new Abstract: Huntington's disease (HD) is a progressive neurodegenerative disorder that affects motor, cognitive, and behavioral functions, where accurate characterization of disease progression remains essential to improve patient outcome and quality of life.

By Lubna Mahmoud Abu Zohair, Hind Zantout

rag safety

arXiv AI

Jun 8

What Your Posts Reveal: A Benchmark and Agentic Framework for User-Level Privacy Leakage on Social Media

arXiv:2606. 06784v1 Announce Type: cross Abstract: Public social media posts can reveal private information through weak cues scattered across text, images, or metadata.

By Zifan Peng, Yini Huang, Aiwen Lu, Qiming Ye, Peixian Zhang, Jingyi Zheng, Yule Liu, Xuechao Wang, Xinlei He, Jiaheng Wei

agents multimodal benchmarks safety

arXiv Machine Learning

Jun 8

Predictable Compression Failures: Order Sensitivity and Information Budgeting for Evidence-Grounded Binary Adjudication

arXiv:2509. 11208v3 Announce Type: replace-cross Abstract: Transformers used for evidence-grounded binary adjudication (e.

By Leon Chlon, Ahmed Karim, Maggie Chlon, MarcAntonio Awada

llms safety

arXiv AI

Jun 8

MVCL-DAF++: Enhancing Multimodal Intent Recognition via Prototype-Aware Contrastive Alignment and Coarse-to-Fine Dynamic Attention Fusion

arXiv:2509. 17446v3 Announce Type: replace-cross Abstract: Multimodal intent recognition (MMIR) suffers from weak semantic grounding and poor robustness under noisy or rare-class conditions.

By Haofeng Huang, Yifei Han, Long Zhang, Bin Li, Yangfan He, Yaxin Xue

multimodal benchmarks safety

arXiv AI

Jun 8

Neuro-Symbolic Learning for Long-Horizon Task Planning Under Complex Logical Constraints

arXiv:2606. 06877v1 Announce Type: cross Abstract: Task planning often suffers from severe efficiency bottlenecks when robots must reason over long-horizon action sequences under complex logical constraints, including object affordances, spatial relationships, and sequential action dependencies.

By Qiwei Du, Zitong Zhan, Shaoshu Su, Bowen Li, Yi Du, Zhipeng Zhao, Taimeng Fu, Sebastian Scherer, Jiaoyang Li, Chen Wang

robotics benchmarks safety

arXiv Machine Learning

Jun 8

TrioPose: Native Triple-Stream Diffusion Transformers for Pose-Guided Text-to-Image Generation

arXiv:2606. 07053v1 Announce Type: cross Abstract: Pose-guided text-to-image generation often suffers from limb distortions and feature crosstalk in complex multi-person scenarios.

By Dian Gu, Zhengyi Yang

llms diffusion multimodal benchmarks safety

arXiv Machine Learning

Jun 8

Elmes*: Automated Construction of Fine-Grained Evaluation Rubrics for Large Language Models in Long-Tail Educational Scenarios

arXiv:2606. 06546v1 Announce Type: new Abstract: Evaluating large language models (LLMs) for education requires measuring how models teach, not only what they know.

By Tao Liu, Ye Lu, Ruohua Zhang, Siyu Song, Wentao Liu, Aimin Zhou, Hao Hao

llms agents benchmarks safety

arXiv AI

Jun 8

SV-Detect: AI-generated Text Detection with Steering Vectors

arXiv:2606. 07313v1 Announce Type: cross Abstract: Detecting machine-generated text is especially difficult under distribution shift, such as transfer across domains, source models, and editing attacks.

By Mikhail Vishnyakov, Tatiana Gaintseva

llms safety

arXiv Machine Learning

Jun 8

Structure-Preserving Correction Learning for Sparse Bayesian Inference in Brain Source Imaging

arXiv:2606. 07196v1 Announce Type: new Abstract: Classical sparse Type-II Bayesian methods for M/EEG brain imaging support joint estimation of source and noise hyperparameters, but rely on fixed iterative update rules.

By Marco Morik, Xiao Ruiting, Shinichi Nakajima, Stefan Haufe, Ismail Huseynov

safety

arXiv AI

Jun 8

A Geometric Account of Activation Steering through Angle-Norm Decomposition

arXiv:2606. 06735v1 Announce Type: new Abstract: Linear activation steering has gained popularity as a simple and empirically effective way to control language model behavior.

By Georgii Aparin, Tatiana Gaintseva

llms safety

Get the daily briefing