AI safety and alignment

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

17,788 stories · RSS feed

arXiv AI

Oct 21

IECG: Weakly-Supervised Interpretability-Enhancing Foundation Model for Electrocardiogram

arXiv:2610.28083v1 Announce Type: new Abstract: Electrocardiography (ECG) is a cornerstone of cardiac diagnosis, yet most machine learning methods for ECG analysis lack interpretability, limiting the...

By Huynh Dang Nguyen, Trong-Thang Pham, Anh-Thuan Do, Phuong-Nhi Dinh, Brijesh Patel, Ngan Le, Van Nguyen

multimodal benchmarks safety

arXiv AI

Oct 21

Fine-Tuning Small Language Models for Reliable VASP INCAR Generation

arXiv:2608.05387v1 Announce Type: cross Abstract: Language models can prepare VASP INCAR files from natural-language requests, but so far only large proprietary cloud models come close to handling th...

By Xinyue Zhang, Jixiang Li, Bin Shao, Baishun Yang, Zhiyang Liu, Weichao Wang

llms fine-tuning efficiency benchmarks safety

arXiv AI

Oct 21

ICLHerd: Herding-Based Data Selection for LLMs with In-Context Learning as a Proxy

arXiv:2610.27142v1 Announce Type: cross Abstract: Data selection plays a central role in modern large language model training, as the quality of training data strongly influences downstream performan...

By Yuqiao Wen, Lili Mou

llms safety

arXiv AI

Oct 21

Target-Aware Curation of Hallucination Stress Tests for Large Language Models

arXiv:2610.27247v1 Announce Type: cross Abstract: Hallucination stress tests can yield similar average scores across language models even when different prompts elicit hallucinations in each model. W...

By Jiho Lee, Bingquan Zhang, Lin Ting Liang, Soroush Vosoughi, Weicheng Ma

llms safety

arXiv AI

Oct 21

Orthogonal Boundary Unlearning for Dual-Use Concepts in Large Language Models

arXiv:2610.27490v1 Announce Type: cross Abstract: Mitigating catastrophic risks from large language models requires excising hazardous operational capabilities, such as cyberwarfare exploitation and...

By Niccol\`o Ronchetti

llms fine-tuning safety

arXiv Computer Vision

Oct 21

Deep learning enables urban change profiling through alignment of historical maps

arXiv:2602.02154v2 Announce Type: replace Abstract: Prior to modern Earth observation technologies, historical maps provide a unique record of long-term urban transformation and offer a lens on the e...

By Sidi Wu, Yizi Chen, Maurizio Gribaudi, Konrad Schindler, Cl\'ement Mallet, Julien Perret, Lorenz Hurni

safety

arXiv AI

Oct 21

Hallucinations of Large Language Models in Cross-Language Code Translation: Phenomena and Mitigation

arXiv:2610.27868v1 Announce Type: cross Abstract: Code translation converts source code from one programming language to another while preserving its functionality. Although Large Language Models (LL...

By Yali Du, Yunji Zhang, Ming Li

llms benchmarks safety

arXiv AI

Oct 21

CountVideoBench: A Benchmark to Evaluate Prediction Biases on the Object Counting task by Video-Language Models

arXiv:2610.28356v1 Announce Type: cross Abstract: Object counting in videos remains a poorly characterized capability of multimodal large language models. Counting questions are included in many exis...

By Marc Chapus, Carlos Crispim-Junior, V\'eronique Eglin, Atilla Baskurt

llms multimodal benchmarks safety

arXiv AI

Oct 21

WorldSR: Harnessing World Knowledge Search for Grounded Image Super-Resolution

arXiv:2610.28423v1 Announce Type: cross Abstract: Generative-prior-based super-resolution (SR) methods leverage implicit knowledge learned from large-scale models to enhance low-resolution inputs. Ho...

By Junyuan Deng, Xinyi Wu, Zhenyao Wu, Song Wang, Yu Wang, Congchao Zhu

benchmarks safety

arXiv AI

Oct 21

The Reward Model Selection Crisis in Personalized Alignment

arXiv:2512.23067v3 Announce Type: replace Abstract: Personalized alignment from preference data has focused primarily on improving personal reward model (RM) accuracy, with the implicit assumption th...

By Fady Rezk, Yuangang Pan, Chuan-Sheng Foo, Xun Xu, Nancy Chen, Henry Gouk, Timothy Hospedales

llms reinforcement-learning fine-tuning benchmarks safety

arXiv AI

Oct 21

Anchoring and Rescaling Attention for Semantically Coherent Inbetweening

arXiv:2603.17651v2 Announce Type: replace-cross Abstract: Generative inbetweening (GI) seeks to synthesize realistic intermediate frames between the first and last keyframes beyond mere interpolation...

By Tae Eun Choi, Sumin Shim, Junhyeok Kim, Seong Jae Hwang

benchmarks safety

arXiv AI

Oct 21

Interpretable PM2.5 Forecasting for Urban Air Quality: A Comparative Study of Adaptive Time-Series Models

arXiv:2603.25495v2 Announce Type: replace-cross Abstract: Accurate short-term air-quality forecasting is essential for public health protection and urban management, yet many recent forecasting frame...

By Moazzam Umer Gondal, Hamad ul Qudous, Asma Ahmad Farhan, Sultan Alamri

safety

arXiv Machine Learning

Oct 21

Privacy Accounting for Empirical Denoisers

arXiv:2610.27710v1 Announce Type: new Abstract: Diffusion models have emerged as the foundation of modern generative systems, yet their high memorization capacity raises privacy concerns. While diffe...

By William Cheng, Mijung Park, Dongjun Kim

diffusion safety

arXiv AI

Oct 21

EDISCO: Equivariant DIScrete Diffusion for Euclidean Combinatorial Optimization

arXiv:2610.04953v2 Announce Type: replace-cross Abstract: Euclidean combinatorial optimization problems (ECOPs), such as the Traveling Salesman Problem (TSP) and Capacitated Vehicle Routing Problem (...

By Ruogu Chen, Jie Han

diffusion benchmarks safety

arXiv AI

Oct 21

Policy Search in Code Space for Multi-Agent Pathfinding

arXiv:2610.25946v2 Announce Type: replace-cross Abstract: Developing coordination policies for complex multi-agent systems relies on either labor-intensive algorithm design or opaque task-specific re...

By Matthias Eder, Pantelis Emmanouil, Vivek Farias, Fabian Fischer, Adam Jozefiak, Tianyi Peng, Thomas Ulz, Raphael Watschinger, Fangyu Wu, Andrew Zheng

agents reinforcement-learning robotics efficiency benchmarks safety

arXiv AI

Oct 21

TamedFlow: Trajectory-Aware Distillation for Teacher-Assisted Diffusion Sampling

arXiv:2610.26025v2 Announce Type: replace-cross Abstract: Diffusion Transformers (DiTs) deliver high-fidelity image and video generation, but inference requires many sequential model evaluations. Two...

By Shuaiqi Cheng, Jiahao Zhang, Boyuan Gu, Yuxi Chen, Haiying Xu, Zexin Lu, Huiyu Zhou, Xuming Hu

llms diffusion efficiency safety

arXiv Machine Learning

Oct 21

A Hybrid TFT-NBEATS Framework for Robust Multi-Horizon Photovoltaic Power Forecasting

arXiv:2610.27155v1 Announce Type: new Abstract: Accurate multi-horizon photovoltaic (PV) power forecasting is essential for grid stability, energy scheduling, and the efficient operation of PVbattery...

By Shafi Bin Sultan, Sabik Bin Sultan, Mohammad Tajrian Arafat Alvi, Rayhanul Islam Sony, Abdulah Al Numan

llms safety

arXiv Machine Learning

Oct 21

Shaped SAEs: Expanding the Toolkit for Mechanistic Interpretability

arXiv:2610.27375v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) seek to decompose the dense activations of large language models (LLMs) into sparse, potentially interpretable features usin...

By Oveys Delafrooz Noroozi, Bhagyashree Puranik, Upamanyu Madhow

llms benchmarks safety

arXiv Machine Learning

Oct 21

From POMDP Theory to Deep RL with Particle Filters

arXiv:2610.27575v1 Announce Type: new Abstract: Model-based reinforcement learning in partially observable environments requires inferring hidden states, learning latent dynamics, and planning under...

By Kevin Tan, Ziping Xu

reinforcement-learning safety

arXiv Machine Learning

Oct 21

Engram Memory Network: Brain-Inspired Prototype Explanations

arXiv:2610.27770v1 Announce Type: new Abstract: Artificial intelligence systems are increasingly deployed in high-stakes decision-making. In response, prototype-based models in explainable AI have at...

By Hyunjun Kim, Myoung Hoon Ha

safety

Get the daily briefing