AI safety and alignment

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

10,089 stories · RSS feed

arXiv AI
Jul 7

dOPSD: On-Policy Self-Distillation for Diffusion Language Models

arXiv:2607. 04428v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) generate text by iteratively denoising a masked sequence, offering a parallel alternative to autoregressive models, but eliciting strong reasoning through post-training remains difficult: supervised fine-tuning is off-policy and suffers from exposure bias, while reinforcement learning gives only sparse, sequence-level rewards and is hard to apply without tractable sequence likelihoods.

By Phuong Tuan Dat, Qi Li, Xinchao Wang
arXiv Machine Learning
Jul 7

F-ACVAE: A Federated Adaptive Conditional Variational Auto-Encoder for Privacy-Preserving Intrusion Detection in IoT Networks

arXiv:2607. 04698v1 Announce Type: new Abstract: The rapid proliferation of Internet of things (IoT) devices has significantly expanded the cyber-attack surface, necessitating robust and privacy-preserving intrusion detection systems (IDS).

By Mohammad Ansarimehr, Somayeh Changiz, Ehsan Baghishani, Ali Mousavi
arXiv Machine Learning
Jul 7

Memisis: Orchestrating and Evaluating Synthetic Data for Tabular Health Datasets

arXiv:2605. 17758v2 Announce Type: replace Abstract: Synthetic data is widely used in healthcare to create datasets that preserve statistical properties of real data without exposing sensitive patient information.

By Nitish Nagesh, Pengbao Zhou, Atchuth Naveen Chilaparasetti, Yajat Nagaraj Kiran, Tu Nguyen, Arshia Harish Puthran, Muhjaazee Love, Aadi Sharma, Mahdi Bagheri, Ian Harris, Amir M. Rahmani
arXiv AI
Jul 7

Hidden-State Privacy Has an Empty Middle

arXiv:2605. 24042v3 Announce Type: replace-cross Abstract: Of $1{,}536$ Gaussian release covariances we tested for single-layer hidden-state privacy, zero achieve both moderate utility and moderate privacy against an adaptive retrieval attacker.

By Alexander Okezue Bell
arXiv AI
Jul 7

From Arithmetic to Logic: The Resilience of Logic and Lookup-Based Neural Networks Under Parameter Bit-Flips

arXiv:2603. 22770v2 Announce Type: replace-cross Abstract: The deployment of deep neural networks (DNNs) in safety-critical edge environments necessitates robustness against hardware-induced bit-flip errors.

By Alan T. L. Bacellar, Sathvik Chemudupati, Shashank Nag, Allison Seigler, Priscila M. V. Lima, Felipe M. G. Fran\c{c}a, Lizy K. John
arXiv AI
Jul 7

Prima.cpp: Fast 30-70B LLM Inference on Heterogeneous and Low-Resource Home Clusters

arXiv:2504. 08791v3 Announce Type: replace-cross Abstract: On-device inference offers privacy, offline use, and instant response, but consumer hardware restricts large language models (LLMs) to low throughput and capability.

By Zonghang Li, Tao Li, Wenjiao Feng, Rongxing Xiao, Jianshu She, Hong Huang, Mohsen Guizani, Hongfang Yu, Qirong Ho, Wei Xiang, Xue Liu
arXiv AI
Jul 7

Test-Time Adaptation for EEG Foundation Models: A Systematic Study under Real-World Distribution Shifts

arXiv:2604. 16926v2 Announce Type: replace-cross Abstract: Electroencephalography (EEG) foundation models have shown strong potential for learning generalizable representations from large-scale neural data, yet their clinical deployment is hindered by distribution shifts across clinical settings, devices, and populations.

By Gabriel Jason Lee, Jathurshan Pradeepkumar, Jimeng Sun