The paper introduces Alignment Forecasting, a method for predicting whether fine‑tuning a language model on a given dataset will increase specific alignment failures such as deception or sycophancy. It presents ALIGNMENTFORECASTBENCH, a benchmark of over 5,000 forecasting questions across many models, datasets, and failure modes, and shows that a simple forecasting scaffold using an LLM’s assessment of dataset bias can outperform baseline forecasters. The authors demonstrate that filtering out high‑risk training examples identified by the forecaster can improve alignment in multiple‑choice evaluations, though benefits in open‑ended conversations remain uncertain.
By Chen Yueh-Han, Bruce W. Lee, Ilia Sucholutsky, Tomek Korbak
Warning: This paper studies stereotypes and biases, and contains potentially disturbing examples, used for illustration purposes only. Our findings should not be interpreted as an argument against alignment.
arXiv:2603. 07445v2 Announce Type: replace-cross Abstract: Large language models (LLMs) often require fine-tuning (FT) to perform well on downstream tasks, but FT can induce safety-alignment drift even when the training dataset contains only benign data.
By Guoli Wang, Haonan Shi, Tu Ouyang, An Wang
arXiv:2607. 03075v1 Announce Type: new Abstract: Safety-critical applications require classifiers that are both robust and reliable.
By Nicolas Sournac, Ahmed Baha Ben Jmaa, Bertrand Braeckeveldt
arXiv:2602. 12124v2 Announce Type: replace Abstract: While most AI alignment research focuses on preventing models from generating explicitly harmful content, a more subtle risk arises from capability-seeking RL training in vulnerable environments.
By Yujun Zhou, Yue Huang, Han Bao, Kehan Guo, Zhenwen Liang, Pin-Yu Chen, Tian Gao, Werner Geyer, Nuno Moniz, Nitesh V Chawla, Xiangliang Zhang
The paper examines whether model uncertainty aligns with human disagreement on vision tasks. Using multi‑annotator datasets (FER+ and CIFAR‑10H), the authors find that pretrained models rarely reflect the ambiguity humans perceive, with weak correlations between model confidence and human disagreement. Predictive multiplicity offers only modest improvement, indicating that common uncertainty metrics fail to flag ambiguous cases.
arXiv:2609.36862v1 Announce Type: cross
Abstract: Fine-tuning-as-a-service lets users adapt a safety-aligned language model to their own data, but it also creates a harmful fine-tuning attack surface...
By Muhammad Zeeshan Akram, Mufid Kamel Marican, Anvesh Reddy Yenugu, Ali Zain Kaimkhani, Minghong Fang
arXiv:2607. 22766v1 Announce Type: cross Abstract: The alignment of Large Language Models (LLMs) is increasingly bottlenecked by data quality.
By Yunting Song, Matthew Watson, Peter Grabowski, Jun Qin
arXiv:2608. 09768v1 Announce Type: new Abstract: A prediction that is both confident and wrong is a critical reliability failure because it can bypass abstention and human review precisely when the model is mistaken.
By Ange-Cl\'ement Akazan, Ineza Remy Mugenga, Abebe Geletu, Jean Medard Ngnotchouye, Issa Karambal
arXiv:2609.37914v1 Announce Type: cross
Abstract: Fine-tuning large language models on narrow, misaligned tasks can undo their post-training alignment and induce novel misaligned behaviors -- a pheno...
By Gon\c{c}alo Paulo, Louis Jaburi, Nora Belrose, Lucia Quirke, Stella Biderman
Fine-tuning-as-a-service lets users adapt a safety-aligned language model to their own data, but it also creates a harmful fine-tuning attack surface: a small amount of harmful data mixed into an othe...
arXiv:2605.01913v2 Announce Type: replace-cross
Abstract: Fine-tuning safety-aligned language models for downstream tasks often leads to substantial degradation of refusal behavior, making models vul...
By Sadia Asif, Mohammad Mohammadi Amiri