AI safety and alignment

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

17,808 stories · RSS feed

arXiv AI

Sep 10

Distributed Lag Neural Additive Models

arXiv:2609.07381v1 Announce Type: cross Abstract: We introduce Distributed Lag Neural Additive Models (DLNAMs), neural-additive analogues of Distributed Lag Non-linear Models (DLNMs) for learning non...

By Calle Helmersson, Shivang Pandey, Leonardo Olivetti, Elena Raffetti

safety

arXiv AI

Sep 10

Same Values, Different Languages? From Multilingual Probing to Steering LLMs Toward Chinese Social Values

arXiv:2609.08515v1 Announce Type: cross Abstract: As Large Language Models (LLMs) are increasingly integrated into human society, aligning them with pluralistic social values has become a critical pr...

By Yuemei Xu, Kexin Xu, Jian Zhou, Haoyu Lu, Yequan Wang, Aishan Liu

llms fine-tuning safety

arXiv AI

Sep 10

Evaluating Steering Techniques using Human Similarity Judgments

arXiv:2505.19333v2 Announce Type: replace Abstract: Current evaluations of Large Language Model (LLM) steering techniques focus on task-specific performance, overlooking how well steered representati...

By Zach Studdiford, Timothy T. Rogers, Siddharth Suresh, Kushin Mukherjee

llms safety

arXiv AI

Sep 10

Large Language Models Transform Organic Synthesis From Reaction Prediction to Automation

arXiv:2508.05427v2 Announce Type: replace Abstract: Large language models (LLMs) are beginning to reshape how organic-synthesis workflows are represented, queried, planned, and connected to experimen...

By Kartar Kumar, Rajesh Kumar, Nikesh Lagun

llms agents robotics multimodal safety

arXiv AI

Sep 10

Behavioral Grammar: Detecting Adaptive Malware via Tiny Language Model Priors and Second-Order Temporal Analysis

arXiv:2608.00745v2 Announce Type: replace Abstract: Modern endpoint detection systems face a fundamental tension: signature-based approaches are trivially evaded by polymorphic or adaptive threats, w...

By Zihan Luo

llms agents safety

arXiv AI

Sep 10

Provable Pluralistic Alignment: Multi-Party RLHF under Offline Human Feedback

arXiv:2403.05006v2 Announce Type: replace-cross Abstract: Pluralistic alignment requires learning from feedback that reflects persistent and potentially conflicting stakeholder preferences while ulti...

By Huiying Zhong, Tianwei Gao, Zhiwei Steven Wu, Linjun Zhang, Weijie J. Su, Zhun Deng

reinforcement-learning safety

arXiv AI

Sep 10

Beyond Retrieval: Joint Supervision and Multimodal Document Ranking for Textbook Question Answering

arXiv:2505.13520v2 Announce Type: replace-cross Abstract: Textbook question answering (TQA) is a complex task, requiring the interpretation of complex multimodal context. Although recent advances hav...

By Hessa Alawwad, Usman Naseem, Areej Alhothali, Ali Alkhathlan, Amani Jamal

llms rag nlp multimodal safety

arXiv AI

Sep 10

BETA-Labeling for Multilingual Dataset Construction in Low-Resource IR

arXiv:2602.14488v3 Announce Type: replace-cross Abstract: IR in low-resource languages remains limited by the scarcity of high-quality, task-specific annotated datasets. Manual annotation is expensiv...

By Md. Najib Hasan, Mst. Jannatun Ferdous Rain, Fyad Mohammed, Nazmul Siddique

llms nlp benchmarks safety

arXiv AI

Sep 10

CTC-TTS: LLM-based dual-streaming text-to-speech with CTC alignment

arXiv:2602.19574v2 Announce Type: replace-cross Abstract: Large-language-model (LLM)-based text-to-speech (TTS) systems can generate natural speech, but most are not designed for low-latency dual-str...

By Hanwen Liu, Saierdaer Yusuyin, Hao Huang, Zhijian Ou

llms rag multimodal safety

arXiv Machine Learning

Sep 10

Structural Entropy-Driven Graph Diffusion Generation for One-Shot Federated Graph Learning

arXiv:2609.06499v1 Announce Type: new Abstract: One-shot federated graph learning (FGL) requires the server to estimate client contributions from highly compressed information, yet conventional volum...

By Shutong Zheng, Lele Fu, Sheng Huang, Wei Yang Bryan Lim, Chuan Chen

diffusion safety

arXiv Machine Learning

Sep 10

Conditioned Initialization for Attention

arXiv:2609.07086v1 Announce Type: new Abstract: Transformers are a dominant architecture in modern machine learning, powering applications across vision, language, and beyond. At the core of their su...

By Hemanth Saratchandran, Simon Lucey

llms safety

arXiv Machine Learning

Sep 10

The Oversight Gap: What LLM Safety Monitors Miss, and Why It Is Not Capability

arXiv:2609.07162v1 Announce Type: new Abstract: Several properties safety monitors are asked to certify, among them cross-tenant noninterference, sandbagging and evaluation awareness, are 2-safety hy...

By Xin Xu

llms benchmarks safety

arXiv Machine Learning

Sep 10

Solving the Elastic Wave Equation with Physics-Informed Neural Networks: A Robust and Critical Assessment

arXiv:2609.07983v1 Announce Type: new Abstract: Physics-Informed Neural Networks (PINNs) have recently emerged as a promising approach for solving Partial Differential Equations (PDEs), offering a me...

By Davide Staub, Ben Moseley

rag safety

arXiv Machine Learning

Sep 10

Miles v0.1: Production-Level Post-Training

arXiv:2609.08368v1 Announce Type: new Abstract: We present Miles v0.1, a full-stack, production-ready system for frontier post-training. Building upon the clean design of slime, Miles designs each st...

By RadixArk, :, Tom Chen, Mao Cheng, Shi Dong, Kangrui Du, Yanbin Jiang, Jiajun Li, Yiming Li, Tao Lin, Yusheng Su, Andy Ye, Yueming Yuan, Zhichen Zeng

agents diffusion reinforcement-learning fine-tuning efficiency safety

arXiv Machine Learning

Sep 10

Characterizing Privacy Risks of Quantum Machine Learning with Emergent Quantum-Native Access

arXiv:2609.05702v2 Announce Type: cross Abstract: Quantum Machine Learning (QML) has shown rapid advances by utilizing quantum computing for machine learning tasks. Meanwhile, the privacy risks accom...

By Liou Tang, James Joshi, Ashish Kundu

safety

arXiv Machine Learning

Sep 10

ModularPhaseNet: Finite-Cyclic Phase Geometry for Computable Semantic Hierarchy, Direction, and Context Consistency in Standard Transformers

arXiv:2609.06000v1 Announce Type: cross Abstract: We propose ModularPhaseNet, a classical and integer-computable discretization of the continuous complex phase geometry introduced in QuantumPhaseNet....

By Kiyotaka Kasubuchi, Kazuo Fukiya

llms efficiency safety

arXiv Machine Learning

Sep 10

Likelihood-Based Unsupervised Anomaly Detection in CMS Dijet Events

arXiv:2609.06686v1 Announce Type: cross Abstract: We present an unsupervised search for anomalous dijet events in proton--proton collision data using neural spline flow density estimation. A normaliz...

By Bhavishya Chebrolu (VIT-AP University, Amaravati, India), Hitesh Rasineni (VIT-AP University, Amaravati, India), Prajwal Aaryan Immadi (VIT-AP University, Amaravati, India)

safety

arXiv Machine Learning

Sep 10

Sub-6 GHz Over-the-Air AMC via Curriculum Fine-Tuned CNN-Transformers

arXiv:2609.07726v1 Announce Type: cross Abstract: Automatic modulation classification (AMC) models are frequently trained and validated on synthetic or channel-cabled data, leaving open the question...

By Nurettin Safak, Muhammet Sefa Demirel, Alperen Marasli, Taha Eren Atmaca, Durdu Can Yerdeyatar, Ozgun Ersoy

llms fine-tuning safety

arXiv Machine Learning

Sep 10

A Review of the Long Horizon Forecasting Problem in Time Series Analysis

arXiv:2506.12809v2 Announce Type: replace Abstract: The long horizon forecasting (LHF) problem has come up in the time series literature for over the last 35 years or so. This review covers aspects o...

By Hans Krupakar, Kandappan V A

efficiency benchmarks safety

arXiv Computation and Language

Sep 10

Dynamics of meaning: Towards the Evaluation of Diachronic Semantic Change in Sinhala

arXiv:2609.08609v2 Announce Type: replace Abstract: Tracking semantic change in low-resource languages across extensive historical timelines presents significant challenges due to data scarcity and t...

By Nevidu Jayatilleke, Nisansa de Silva

llms rag fine-tuning efficiency safety

Get the daily briefing