paper-with-me

홈 › Papers

Steering LLMs via Scalable Interactive Oversight

2026-02-04 · Enyu Zhou, Zhiheng Xi, Long Ma, Zhihao Zhang, Shihan Dou, Zhikai Lei, Guoteng Wang, Rui Zheng, Hang Yan, Tao Gui, Qi Zhang, Xuanjing Huang arxiv

As Large Language Models increasingly automate complex, long-horizon tasks such as \emph{vibe coding}, a supervision gap has emerged. While models excel at execution, users often struggle to guide them effectively due to insufficient domain expertise, the difficulty of articulating precise intent, and the inability to reliably validate complex outputs. It presents a critical challenge in scalable oversight: enabling humans to responsibly steer AI systems on tasks that surpass their own ability to specify or verify. To tackle this, we propose Scalable Interactive Oversight, a framework that decomposes complex intent into a recursive tree of manageable decisions to amplify human supervision. Rather than relying on open-ended prompting, our system elicits low-burden feedback at each node and recursively aggregates these signals into precise global guidance. Validated in web development task, our framework enables non-experts to produce expert-level Product Requirement Documents, achieving a 54\% improvement in alignment. Crucially, we demonstrate that this framework can be optimized via Reinforcement Learning using only online user feedback, offering a practical pathway for maintaining human control as AI scales.

📄 PDF Abstract BibTeX arXiv:2602.04210

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Vis-CoT: A Human-in-the-Loop Framework for Interactive Visualization and Intervention in LLM Chain-of-Thought Reasoning

2025-09-01 · Kaviraj Pather, Elena Hadjigeorgiou, Arben Krasniqi, Claire Schmit 외 arxiv

Large language models (LLMs) show strong reasoning via chain-of-thought (CoT) prompting, but the process is opaque, which makes verification, debugging, and control difficult in high-stakes settings. We present Vis-CoT, …

AdaLens: Interactive Storyline for Monitoring and Steering Long-Running Agentic Data Analysis

2026-08-18 · Yangtian Liu, Yan Miao, Shuhan Liu, Yunfan Zhou 외 arxiv

Large language models are pushing data science toward increasingly autonomous and agentic workflows, with recent systems already supporting multi-step and long-running analyses. As these workflows become more autonomous,…

LLM Wardens: Mitigating Adversarial Persuasion with Third-Party Conversational Oversight

2026-05-08 · Lennart Wachowiak, Scott D. Blain, David Williams-King, Samuele Marro arxiv

LLMs are increasingly capable of persuasion, which raises the question of how to protect users against manipulation. In a preregistered user study (N=120) across four decision-making scenarios, we find that an adversaria…

Negative Before Positive: Asymmetric Valence Processing in Large Language Models

2026-05-07 · Sohan Venkatesh arxiv

Mechanistic interpretability has revealed how concepts are encoded in large language models (LLMs), but emotional content remains poorly understood at the mechanistic level. We study whether LLMs process emotional valenc…

Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges

2026-04-15 · Xiaohua Wang, Muzhao Tian, Yuqi Zeng, Zisu Huang 외 arxiv

Reinforcement Learning from Human Feedback (RLHF) and related alignment paradigms have become central to steering large language models (LLMs) and multimodal large language models (MLLMs) toward human-preferred behaviors…

Reinforcement Learning