The paper investigates how autoresearch agents—AI systems that conduct research autonomously—benefit from interacting with human researchers, a process called co‑improvement. It presents empirical evidence that humans remain crucial during the ideation phase of benchmark creation, introduces HackBench to expose alignment gaps in current agents, and proposes two training methods: using human simulators to enhance agent interaction and applying reinforcement learning with expert‑aligned rubrics to improve communication. These approaches aim to make autoresearch agents more capable and aligned with human expectations.
By Tianjian Li, Chuanyang Jin, Seungone Kim, Tong Chen, John Hoffman, Yoram Bachrach amd Pedro Rodriguez, Ping Yu, Eryk Helenowski, Swarnadeep Saha, Jack Lanchantin, Weizhe Yuan, Ilia Kulikov, Jason Weston
The paper introduces Test-Time Adaptation through Human‑Agent Interaction (TAHI), a method that leverages iterative human feedback to tailor AI agents to individual users. By integrating interaction data into agent context and weighting, and by developing an evolving rubric module, the authors adapt agents to 30 users across writing and visual creation tasks, achieving 4.5–20.9% higher solo task success after only a few interactions. The rubric module also serves as a scalable annotation tool, detecting 16.0–22.3% more failures than language models or humans alone, while personalized agents can generalize improvements up to 8.8% across users.
By Zora Zhiruo Wang, Apurva Gandhi, Rulin Shao, Aspen Chen, Jonas Mueller, Zhiqi Liang, Jett Chen, Michael Ryan, Qianou Ma, Luxi He, Zhoujun Cheng, Andre He, Seungone Kim, Jiayi Geng, Mingqian Zheng, Weiwei Sun, Zheyuan Zhang, Xinran Zhao, Yike Wang, Abe Hou, Liwei Jiang, Pang Wei Koh, Diyi Yang, Graham Neubig, Daniel Fried
The paper introduces Test-Time Adaptation through Human‑Agent Interaction (TAHI), a method that uses iterative human feedback to adapt AI agents to individual users’ criteria. By integrating cross‑session interaction data into agent context and weights, and building an evolving rubric module, the authors demonstrate that agents can improve task success by 4.5–20.9% after only a few interactions. The evolving rubric also serves as a scalable annotation tool, detecting 16.0–22.3% more failures than language models or humans alone, and personalized agents can even generalize improvements up to 8.8% across users.
By Zora Zhiruo Wang, Apurva Gandhi, Rulin Shao, Aspen Chen, Jonas Mueller, Zhiqi Liang, Jett Chen, Michael Ryan, Qianou Ma, Luxi He, Zhoujun Cheng, Andre He, Seungone Kim, Jiayi Geng, Mingqian Zheng, Weiwei Sun, Zheyuan Zhang, Xinran Zhao, Yike Wang, Abe Hou, Liwei Jiang, Pang Wei Koh, Diyi Yang, Graham Neubig, Daniel Fried
arXiv:2607. 16610v1 Announce Type: new Abstract: Long-horizon AI agents are becoming increasingly capable, yet their interaction with users remains surprisingly thin.
By Chen Chen, Zhehuai Chen
arXiv:2608. 14870v1 Announce Type: new Abstract: Long-horizon agents can execute continuously, but human attention remains intermittent and scarce.
By Chen Chen, Zhehuai Chen
arXiv:2608. 01366v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are integral to complex intellectual tasks, yet output quality remains constrained by user-provided prompts.
By B. Sankar, Pawni Yadav, Srinidhi Ranjini Girish, Amogh A. S
arXiv:2607. 20773v1 Announce Type: cross Abstract: Large language models (LLMs) have shifted human--computer interaction from `traditional'' interface journeys toward more conversational exchanges.
By Zeshu Zhu, Natalie Friedman, Kevin Weatherwax, Emily Eiben
arXiv:2602. 11351v2 Announce Type: replace Abstract: Proactive large language model (LLM) agents aim to actively plan, query, and interact over multiple turns, enabling efficient task completion beyond passive instruction following and making them essential for real-world, user-centric applications.
By Yihang Yao, Zhepeng Cen, Haohong Lin, Shiqi Liu, Zuxin Liu, Jiacheng Zhu, Zhang-Wei Hong, Laixi Shi, Ding Zhao
arXiv:2609.16436v1 Announce Type: cross
Abstract: Simulations based on large language models (LLMs) have proven to be powerful for understanding human behavior, making them valuable additions to the...
By Jiayue Gaveal Fan, Arul Murugan, Shreyas Krishnan, Abhishek Nagaraj
The paper introduces A-B-D, a method that infers personality traits of large language model agents from their behavioral data, combining functional actions and linguistic outputs across 345,667 real-world trajectories of 80 models. After filtering for stability and discriminability, 79 features are reduced to six reliable factors—two functional and four linguistic—revealing differences such as Claude‑Sonnet‑4.6’s planfulness versus GPT‑5.5’s low energy. The study also demonstrates a weak correlation between these inferred traits and agents’ self-reported Big Five scores, highlighting a knowledge‑action gap.
By Haokai Zhao, Jie Gao, Yunze Xiao, Xintao Wang, Weihao Xuan, Aditya Joshi, Mark Dredze, Jen-tse Huang
The paper introduces pluralistic agents—AI representatives that deliberate on behalf of individuals—to scale public input on AI policy. Experiments with 746 participants show that these agents can reliably shift policies toward their represented personas’ preferences and are viewed as useful and fair, outperforming a single AI reasoning model. The study highlights how agent influence varies with instructions and design, suggesting a cautious but promising role for pluralistic agents in expanding human participation in AI safety policy development.
By Jacy Reese Anthis, Ken Burke, David Madras, Crystal Qian, Meredith Ringel Morris, Logan Cross, Caroline Pantofaru, Utsav Prabhu
SocialArena is a benchmark designed to evaluate the social intelligence of large language models (LLMs) across three dimensions: role‑playing ability, mental state reasoning, and behavioral reasoning. It introduces a role‑playing dialogue paradigm with explicit character thinking to assess these capabilities in realistic human interactions, and a U‑TRIMF framework to synthesize diverse social contexts. The evaluation relies on a human‑annotated arena‑style protocol, supported by a $132.4K investment in recruiting and training annotators who compare model responses and provide preference labels and critiques.
By Zongqi Wang, Rui Wang, Yuchuan Wu, Hua Yang, Bingxue Qiu, Tianle Gu, Bing Zhao, Yujiu Yang, Jieping Ye, Yongbin Li