The paper introduces COLAGUARD, a guardrail model that embeds multi-step safety reasoning into a continuous latent space, allowing efficient hidden-state propagation during inference. Compared to existing methods, COLAGUARD achieves an 8.24‑point macro‑F1 improvement over Llama Guard 3 and matches the explicit reasoning baseline GuardReasoner, while delivering a 12.9× speedup and a 22.4× reduction in token usage across ten moderation settings and eight safety benchmarks.
By Siddharth Sai, Xiaofei Wen, Muhao Chen
The paper introduces SimpleLoRIF, a scalable gradient‑embedding technique that pinpoints fine‑tuning examples whose gradients align with unsafe behavior in large language models. By applying this method, the authors create SafeTulu, a version of Tulu 3 with the 20 % of most unsafe‑aligned examples removed, which reduces GCG attack success by up to 22 % across several LLMs while largely preserving or improving capability. The study also highlights previously under‑studied safety‑degrading data sources such as tool‑execution failures and narrative arithmetic.
By Saket Reddy, Amro Abbas, Anshuman Suri, Ari Morcos, David Schwab, Pratyush Maini
arXiv:2604. 07223v2 Announce Type: replace-cross Abstract: As large language models (LLMs) evolve from static chatbots into autonomous agents, the primary vulnerability surface shifts from final outputs to intermediate execution traces.
By Yen-Shan Chen, Sian-Yao Huang, Cheng-Lin Yang, Yun-Nung Chen
AdaGuard is an adaptive guard model that evaluates language‑model agent trajectories against user‑defined policies. The authors created the AdaptiveSafety dataset, containing 10,939 training and 1,000 test examples with policy rules, counterfactuals, explanations, and violated rules. Using supervised pre‑training followed by the SafePO reinforcement‑learning algorithm, they built 0.6B, 4B, and 8B guard models, with the 4B version achieving 89.30% accuracy on AdaptiveSafety and 71.82% on DynaBench.
By Yunhao Feng, Yifan Ding, Yuxiang Xie, Zheng Li, Mingrui Lao, Zeyuan Wang, Yanming Guo
arXiv:2607. 11475v1 Announce Type: new Abstract: Safety alignment in large language models can be fragile under fine-tuning, as even benign task adaptation may increase harmful compliance.
By Aznaur Aliev, Carlos Hinojosa, Abdelrahman Eldesokey, Bang An, Bernard Ghanem, Yibo Yang
arXiv:2610.31713v1 Announce Type: cross
Abstract: Advances in Omni-modal Large Language Models (OLLMs) have enabled increasingly seamless human-AI interaction, but their unified processing of heterog...
By Junhao Wang, Yidou Chen, Li Lu, Hongwei Yao, Xiaodan Li, Mengjie Wu, Yuefeng Chen, Zhan Qin, Kui Ren
Safety alignment in large language models can be fragile under fine-tuning, as even benign task adaptation may increase harmful compliance. Existing defenses mainly follow two directions: they either intervene during or after fine-tuning through retraining or weight modification, which can be costly and may hurt task performance, or they use model-agnostic safety classifiers, which may miss failures specific to a given fine-tuned checkpoint.
Vision-language models (VLMs) are increasingly deployed in consumer, medical, financial, and enterprise applications. This broad deployment expands the safety surface: risks can arise from multimodal question answering, assistant responses, and cross-modal composition, while moderation policies may vary across products, regions, and deployment stages.
arXiv:2607. 02072v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed in domains requiring guardrails to detect unsafe, off-topic, or adversarial prompts.
By Mahmoud Abdelfattah, Hamid Nasiri, Peter Garraghan
Guardrail models, which screen malicious prompts in LLM services, often use lightweight Transformers with short context windows and bucketed positional encodings. The study identifies a new failure mode called Overflip, where repeating a prompt causes the guardrail’s prediction to flip from malicious to benign as the sequence length increases. Experiments on nine popular guardrails show that 5 models exhibit MAL→BEN flips on 100 prompts, with flip rates ranging from 8% to 92% and first flips occurring between 2.6k and 9.4k tokens, highlighting a gradual attention dispersion distinct from traditional attention‑dilution attacks.
By Xu He, Chih-Hsuan Lin, Hung-Mao Chen, Junjie Xiong, Yan Zhai, Kun Sun
arXiv:2606. 05805v1 Announce Type: new Abstract: LLM-based guardrails typically safeguard agents by evaluating proposed actions or inputs before execution, producing safety signals such as binary allow/deny decisions, risk categories, and/or explanatory rationales about potential policy violations.
By Yuhao Sun, Jiacheng Zhang, Shaanan Cohney, Zhexin Zhang, Feng Liu, Xingliang Yuan
arXiv:2606. 29887v1 Announce Type: new Abstract: In real-world applications, guardrails are often expected to identify unsafe user-model interactions according to application-specific safety policies, rather than relying on predefined risk taxonomies.
By Jiacheng Zhang, Haoyu He, Sen Zhang, Shen Wang, Xiaolei Xu, Yuhao Sun, Meng Shen, Feng Liu