The paper introduces PAIR, a pipeline that creates user-conditioned application states to evaluate mobile GUI agents across different users, and RePAIR, a reinforcement learning method that incorporates personalization-aware interaction rewards. Experiments with six agents reveal significant variability in task success and subgoal achievement across users, especially when targeting personal content. RePAIR improves success rates over supervised fine-tuning, suggesting that learning from cross-user differences can enhance GUI-agent reliability.
By Yeji Park, Jaeyun Shim, Taesik Gong
SIGMA is a self‑improving alignment framework that uses a model’s own reasoning to generate diverse alignment dilemma scenarios from a simple "Model Spec" and then fine‑tunes the model with self‑judged rewards. By performing spec‑guided task synthesis and rubric‑based reinforcement learning, SIGMA improves safety alignment in multi‑turn agentic settings, reducing harmfulness and misalignment scores while preserving general capability. The approach demonstrates that a model can autonomously strengthen its own safety reasoning without external supervision.
By Jingyu Zhang, Shruti Palaskar, Daniel Khashabi, Benjamin Van Durme, Leon A. Gatys, Joseph Yitan Cheng
POLAR is a guardrail framework designed for small tool‑calling language‑model agents that evaluates the reversibility of each action using a two‑layer ontology. By assigning graded reversibility scores and pruning calls that fall below a threshold, POLAR aims to prevent operational risks before they occur. In experiments on the τ²‑bench across six agent models, POLAR increased mean task reward by 0.11 to 0.18 points on airline tasks for four of six agents, though improvements were limited to eight of eighteen model‑domain cells, with some domains showing regression.
By Yunju Kang, Seonghyeon Cho, Irene Li, Yeo-Chan Yoon, Chanjun Park
The paper introduces SSRFT (Supervised Safe-Role Fine‑Tuning), a new framework that reframes safety alignment for large language models as internalizing a predefined safe role rather than relying on explicit refusal patterns. SSRFT builds a Safe‑Role Question‑Answer dataset from psychometric questions, limited jailbreak prompts, and a safe‑role description, then synthesizes and validates role‑consistent responses to expose models to diverse safety scenarios. Experiments on multiple base and instruct models demonstrate that SSRFT yields more robust and generalizable safety alignment, better resilience to prefilling attacks, improved generalization to unseen jailbreak domains, and reduced over‑refusal on benign queries while preserving general capabilities.
By Jinghao Pang, Jitai Hao, Qiang Huang, Zhaochun Ren, Jun Yu
The paper investigates when adding a second AI component—an AI supervisor—benefits network‑fraud decision making. Using a role‑aware Decider‑Supervisor framework with blockchain auditability, the authors evaluate four configurations combining centralized ML, federated meta‑models, and large language model variants. Results show that only specific setups (e.g., DS‑4 Base in extreme high‑fraud scenarios) reduce errors, while most supervised configurations either underperform or add operational burden.
By Saviz Changizi, Nasibeh Mohammadzadeh, Mohammad Shojafar, Rahim Tafazolli
arXiv:2610.07560v1 Announce Type: new
Abstract: Goal-directed molecular design has advanced rapidly, yet a substantial proportion of designed molecules remain difficult to synthesize in practice, lim...
By Tao Li, Tuan Vinh, Monika Raj, Yuan Fang, Zhichun Guo, Carl Yang
arXiv:2610.06994v1 Announce Type: cross
Abstract: Backdoor-defense leaderboards print a clean-accuracy drop and read it as removal cost. Measured on the poisoned victim alone, the drop cannot separat...
By Ruizhi Xu, Wei Xu, Sibo Zhu
arXiv:2610.07032v1 Announce Type: cross
Abstract: Large-scale pretrained transformer models have achieved state-of-the-art performance across diverse machine translation tasks, including multilingual...
By Maria Zafar, Souhail Bakkali, Rejwanul Haque
arXiv:2610.07338v1 Announce Type: cross
Abstract: Audio benchmarks are built around short, pre-segmented clips, limiting model design to brief inputs or fixed vocabularies. To close this gap, we intr...
By Kwanghee Choi, Suwon Shon, Dmitriy Serdyuk, Guitang Lan, Chao-Wei Huang, Mohammad Sadegh Rasooli, Sangeeta Srivastava, Zhaojiang Lin, Saurabh Adya, Ming Sun
arXiv:2610.07518v1 Announce Type: cross
Abstract: Safety auditing of post-trained large language models typically relies on model behavior, requiring model execution and depending on the coverage of...
By Ziqun Bao, Xinyu Zhang, Yuchen Shao, Chengcheng Wan
arXiv:2610.07553v1 Announce Type: cross
Abstract: LoRA fine-tuning adapts small language models (SLMs) to heterogeneous instruction data within a low-rank update subspace, making it vulnerable to thr...
By Hongyu Cao, Yanchi Liu, Kunpeng Liu, Xujiang Zhao, Wei Cheng, Zhengzhang Chen, Yanjie Fu, Haifeng Chen
arXiv:2610.07603v1 Announce Type: cross
Abstract: This article proposes Selective Affective Layer Fine-Tuning (SALFT), an efficient adaptation framework for Video Vision Transformers in player arousa...
By Yi Xia, Ibrahim Khan, Mury Fajar Dewantoro, Wenwen Ouyang, Ruck Thawonmas
arXiv:2610.07764v1 Announce Type: cross
Abstract: Natural language processing (NLP) models can detect depression-related language in text written near the time symptoms are measured, but whether pret...
By Daniel Kua, Emrul Hasan, John-Jose Nunez, Frances Chen
arXiv:2610.08123v1 Announce Type: cross
Abstract: End-to-end planners based on waypoint regression achieve strong open-loop accuracy, but they primarily learn to mimic expert geometry and remain diff...
By Ahmed Abouelazm, Rupert Polley, Qingyuan Zhang, Yin Wu, Philip Sch\"orner, Carl Esselborn, J. Marius Z\"ollner
arXiv:2610.08161v1 Announce Type: cross
Abstract: Ambient documentation systems are rapidly gaining adoption, yet their impact on clinical note quality remains poorly characterized. We introduce MedC...
By Daniel Varab, Victor Petr\'en Bach Hansen, Asbj{\o}rn W. Helge, Kevin Pelgrims, Mathias Baltzersen, Adrian Young-San Roessler, Vanessa Klungtvedt, Maximilian Brand, Lasse Krogsb{\o}ll, Henrik Cullen, Lars Maal{\o}e
arXiv:2610.08773v1 Announce Type: cross
Abstract: Web agents complete user requests by reading and acting on pages that third parties write, so an instruction planted on a page can redirect the agent...
By Sarim Hashmi, Mukul Ranjan, Kshitij Mishra, Mikhail Kuznetsov, Praneeth Vepakomma, Nils Lukas
arXiv:2605.06183v2 Announce Type: replace
Abstract: Low-rank adaptation (LoRA) is a widely used parameter-efficient fine-tuning method that places trainable low-rank adapters into frozen pre-trained...
By Suoxin Zhang, Run He, Di Fang, Xiang Tan, Kaixuan Chen, Huiping Zhuang
arXiv:2410.21582v4 Announce Type: replace-cross
Abstract: Large-scale pretrained models are widely leveraged as foundations for learning new specialized tasks via fine-tuning, with the goal of mainta...
By Jaedong Hwang, Brian Cheung, Zhang-Wei Hong, Akhilan Boopathy, Pulkit Agrawal, Ila Fiete
arXiv:2511.21075v4 Announce Type: replace-cross
Abstract: Engineering LLMs to accelerate life sciences research requires a robust alignment with biomedical knowledge. We observe that biomedical text...
By Zhenchao Tang, Fang Wang, Haohuai He, Jiale Zhou, Tianxu Lv, Jun Zhu, Shouzhi Chen, Minghao Yang, Yu Wang, Jiayang Wu, Yidong Song, Yaokun Li, Jiehui Huang, Jun Zhou, Bing He, Jianhua Yao
arXiv:2603.02655v2 Announce Type: replace-cross
Abstract: Real-time video commentary generation provides textual descriptions of ongoing events in videos. It supports accessibility and engagement in...
By Anum Afzal, Yuki Saito, Hiroya Takamura, Katsuhito Sudoh, Shinnosuke Takamichi, Graham Neubig, Florian Matthes, Tatsuya Ishigaki