cs.AI · 2026-07-16 · No. 55

Artificial Intelligence, 2026-07-16.

32 new papers in cs.AI. Titles, authors, abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →

01 — The papers

32 entries
  1. 01

    Deep Interaction: An Efficient Human-AI Interaction Method for Large Reasoning Models

    Hefeng Zhou, Jinxuan Zhang, Jiong Lou, Yuxin Liu, Chaochao Lu, Jingjing Qu, Jie Li

    cs.AI

    The emergence of Chain-of-Thought (CoT) reasoning has significantly enhanced the ability of large language models (LLMs) to tackle complex, multi-step tasks. However, when errors occur, current interaction approaches typically involve re-generating another response that may make mistakes again, or users laboriously flag the faulty step in follow-up turns that may get responses <You are right, I made a mistake here> followed by similar errors...

    arxiv.org/abs/2607.14049 · PDF

  2. 02

    Earthquaker-AI: A Retrieval-Augmented Generation Framework with Rubric-Based Assessment for Primary School Earthquake Education

    Xanthi Kokkinou, Chaido Mizeli, Nafsika Koulaxidou, Marina Delianidi, Konstantinos Diamantaras

    cs.AI

    This paper presents Earthquaker-AI, a hybrid educational framework building upon a previously implemented educational robotics project by integrating a conversational AI assistant based on Retrieval-Augmented Generation. It aims to enhance earthquake preparedness and conscious action among primary-school students. The system extends the award-winning STEM project Earthquaker moving from mechanical simulation with Lego WeDo2 to cognitive and...

    arxiv.org/abs/2607.14046 · PDF

  3. 03

    AI-accelerated End-to-End Framework for Rapid Professional Upskilling

    Tam Nguyen, Hung Nguyen, Robert Ogburn

    cs.AI

    By 2030, 59 of every 100 workers will need reskilling or upskilling, yet the average time to close an enterprise skills gap grew from roughly 3 days in 2014 to 36 days in 2018. Most current frameworks accelerate single stages of upskilling programs and generally lack industry validation. We present an end-to-end framework that applies AI acceleration across five stages of knowledge acquisition, content development, content review and...

    arxiv.org/abs/2607.14044 · PDF

  4. 04

    Do Agent Optimizers Compound? A Continual-Learning Evaluation on Terminal-Bench 2.0

    Wenxiao Wang, Priyatham Kattakinda, Soheil Feizi

    cs.AI · cs.CL · cs.LG

    Most reported gains from agent-optimization methods are one-shot: an agent is optimized against a fixed benchmark and the resulting improvement is reported as if it were a stable property of the method. This does not test the setting that matters for deployed agents, where optimization is applied recursively as new failures and new tasks appear over time. The central question this raises is whether optimizer-driven gains compound: after an...

    arxiv.org/abs/2607.14004 · PDF

  5. 05

    A Self-Evolving Agent for Longitudinal Personal Health Management

    Haoran Li, Jiebi Deng, Tong Jin, Jinghong Han, Yuxin Wang, Zexin Wang, Qingyi Si, Weikang Gong, Xiahai Zhuang, Jia...

    cs.AI

    Personal health management unfolds over repeated encounters, yet most health AI systems treat each request in isolation. We developed HealthClaw, an open-source agent architecture that updates support as a person's routines, preferences, measurements and risks change. It separates shared safety rules and medical knowledge from private longitudinal memory containing profile facts, reusable procedures and episodic traces. After each episode,...

    arxiv.org/abs/2607.13940 · PDF

  6. 06

    AIMO Interpretability Challenge

    Michal Štefánik, Philipp Mondorf, Andreas Waldis, Qianying Liu, Chuan Yang, Michal Spiegel, Josef Kuchař, Marek...

    cs.AI

    We propose the AIMO Interpretability Challenge, a competition on distinguishing robust from spurious reasoning in frontier mathematical language models based on the models' internal mechanisms. The challenge is motivated by a central limitation of standard reasoning benchmarks: strong final-answer accuracy does not reveal whether a model relies on stable reasoning mechanisms or exploits brittle reasoning shortcuts. Building on AI Mathematical...

    arxiv.org/abs/2607.13899 · PDF

  7. 07

    Experience Memory Graph: One-Shot Error Correction for Agents

    Wenjun Wang, Yuchen Fang, Fengrui Liu, Zibo Liang, Kai Zheng

    cs.AI

    Large Language Model (LLM) agents have shown remarkable capabilities in autonomous decision-making by generating sequential trajectories of states, actions, and observations. However, in complex, long-horizon tasks, these agents frequently suffer from compounding errors and struggle to recover from failures. Existing self-correction mechanisms rely on prompt-based reflection, which is inherently brittle, incurs heavy time and API costs due to...

    arxiv.org/abs/2607.13884 · PDF

  8. 08

    CAVA: Canonical Action Verification and Attestation for Runtime Governance of Agentic AI Systems

    Zexun Wang

    cs.AI

    Agentic AI systems increasingly act through heterogeneous runtimes: local coding hooks, SDK tools, browser automation, managed-agent traces, API gateways, and workflow engines. A single operational act such as publishing code, changing identity state, moving money, or exporting data may therefore be represented by many incompatible runtime records. This makes a basic governance question difficult to answer: what action was actually approved,...

    arxiv.org/abs/2607.13716 · PDF

  9. 09

    AgentCompass: A Unified Evaluation Infrastructure for Agent Capabilities

    Zichen Ding, Jiaye Ge, Shufan Jiang, Kai Chen, Mo Li, Qingqiu Li, Zehao Li, Zonglin Li, Tiaohao Liang, Shudong Liu,...

    cs.AI · cs.SE

    As Large Language Models (LLMs) evolve into autonomous agents, the need for unified evaluation infrastructure becomes critical. However, current evaluation pipelines remain highly fragmented and tightly coupled, hindering reproducibility and causing redundant engineering. To address this, we introduce AgentCompass, an open-source, lightweight, and extensible infrastructure for evaluating LLM-based agents. AgentCompass organizes the evaluation...

    arxiv.org/abs/2607.13705 · PDF

  10. 10

    When Bots Join the Team: Bot Adoption and the Institutional Fabric of Open-Source Software Projects

    Yongren Shi, Wenyi Gong

    cs.AI · cs.HC

    AI agents are joining human teams, raising a basic question: when an automated agent becomes a regular participant, does group organization strengthen or weaken? We study this question in open-source software, where bots open pull requests, review code, and merge changes alongside people, leaving a public record of every interaction. Treating bots as participants rather than tools, we examine 2,991 GitHub projects for two years before and...

    arxiv.org/abs/2607.13679 · PDF

  11. 11

    Explaining Reinforcement Learning Agents via Inductive Logic Programming

    Celeste Veronese, Edoardo Zorzi, Daniele Meli, Alessandro Farinelli

    cs.AI

    Explainable Reinforcement Learning (XRL) seeks to make Reinforcement Learning (RL) policies more transparent and interpretable, a key requirement in safety-critical and human-centric scenarios. However, it is mostly based on user studies, thus targeting the needs of a specific audience and lacking shared evaluation metrics. On the other hand, logic-based approaches within eXplainable Artificial Intelligence (XAI) provide compact,...

    arxiv.org/abs/2607.13655 · PDF

  12. 12

    UESF-Bench: Benchmarking and Probing for Unified Embodied Seeking and Following

    Kun Yu, Jianhua Yang, Yixiang Chen, Changwei Wang, Hongyuan Yu, Yan Huang, Fushuo Huo, Ya Jing, Zhumin Chen, Keji He

    cs.AI

    Language-guided human following is an important capability for embodied agents, but existing benchmarks typically assume that the target person is visible at the start of an episode. This setting simplifies the problem and overlooks a more realistic requirement: an agent often needs to first find a language-described target and then persistently follow that target in a dynamic environment. While recent work has started to study human search,...

    arxiv.org/abs/2607.13621 · PDF

  13. 13

    STOCKTAKE: Measuring the Gap Between Perception and Action in LLM Agents with a Fair Oracle

    Sagar Deb, Ashwanth Krishnan

    cs.AI · cs.LG

    LLM agents are increasingly evaluated on multi-week decision tasks in which the state that drives cost is never directly observed. On such tasks the final cost cannot say why an agent failed: it may have misread the world, or read it correctly and still failed to act (the knowing-doing gap). Existing evaluations cannot separate these two failures; their reference policies either read privileged information the agent never sees, or are missing...

    arxiv.org/abs/2607.13618 · PDF

  14. 14

    Automatic Ordinary Differential Equations Discovery For Biological Systems Using Large Language Model Powered Agentic System

    David Krongauz, Arad Zulti, Eran Segal, Teddy Lazebnik

    cs.AI · math.DS

    Automatic scientific discovery has long been a goal of computational scholars - a machine that can discover nature's secrets on its own, moving computational systems beyond data-fitting tools toward the generation and refinement of mechanistic models of the universe. Recent advances in symbolic regression (SR) and large-language-model (LLM)-based agents suggest that such systems can recover equations from data, incorporate domain priors, and...

    arxiv.org/abs/2607.13608 · PDF

  15. 15

    SAFETY SENTRY: Context-Aware Human Intervention via EXECUTE-ASK-REFUSE Routing

    Tianyu Chen, Chujia Hu, Wenjie Wang

    cs.AI

    LLM agents act on real-world environments through tool calls, and a single misjudged action can cause irreversible harm. The standard safeguard is a guard model that labels each proposed action as safe or unsafe, but this binary view conflates two distinct decisions: whether the action is harmful in itself, and whether it is appropriate given the user's context. It also operates at the granularity of action categories rather than individual...

    arxiv.org/abs/2607.13594 · PDF

  16. 16

    AI advice suppresses people's willingness to say "I don't know", even when the advice is wrong and accuracy is incentivized

    Chiara Marcoccia, Walter Quattrociocchi, Valerio Capraro

    cs.AI · cs.CY · cs.HC

    Knowing when to say "I don't know" is fundamental to human judgment, yet AI assistants offer a fluent answer to almost any question. In five experiments (N = 3,132; four preregistered, one direct replication), participants answered difficult questions and could always decline to respond. We engineered the questions so that AI advice was wrong, separating AI use from its accuracy. Merely having access to AI nearly eliminated participants'...

    arxiv.org/abs/2607.13562 · PDF

  17. 17

    Multi-Agent Collaborative Reasoning with Tool-Augmented Evidence for Urban Region Profiling

    Xixuan Hao, Yutian Jiang, Jiabo Liu, Yihang Yang, Guangyin Jin, Song Gao, Yuxuan Liang

    cs.AI

    Urban region profiling constitutes a core problem in urban computing, supporting applications such as population estimation, economic assessment, and environmental monitoring. Existing methods typically formulate this task as multimodal representation learning, fusing heterogeneous urban data, e.g., satellite imagery, points of interest, textual descriptions, and 3D building information, into latent embeddings for prediction. However, these...

    arxiv.org/abs/2607.13558 · PDF

  18. 18

    How Far Can Root Cause Analysis Go on Real-World Telemetry Data?

    Athira Gopal, Ashwanth Krishnan

    cs.AI

    Identifying root causes in production microservice failures requires reasoning over large-scale, multimodal telemetry spanning metrics, logs, and traces, a problem that has proved resistant to both classical and LLM-based approaches. The OpenRCA dataset exemplifies these challenges: it is large-scale, multimodal, and lacks detailed domain knowledge, and yields consistently low accuracy across all existing methods. We show that classical...

    arxiv.org/abs/2607.13548 · PDF

  19. 19

    LAPO: Leave-One-Turn Attribution for Self-Generated Process Rewards in Multi-Turn Search Reasoning

    Qiang Zhu, Jiajun Wu

    cs.AI

    Reinforcement learning for multi-turn search reasoning typically relies on terminal outcome rewards, which cannot distinguish useful, redundant, and harmful intermediate interactions. We propose LAPO, a self-generated process-supervision method based on backward leave-one-turn attribution. For each search turn, LAPO replaces the turn and its retrieval observation with a fixed [DELETE] placeholder and measures the resulting change in the...

    arxiv.org/abs/2607.13501 · PDF

  20. 20

    Set-shifting Behavioral Test for Harnessed Agents

    Ziwei Ye

    cs.AI · cs.CL · cs.SE

    What happens to an LLM agent's tool choice when the reliable tool silently changes within an ongoing session? We borrow set-shifting from cognitive psychology to study how well agents adapt to hidden reliability shifts. Our benchmark mounts tool-skill libraries with redundancies, where many tools solve the same task but differ in hidden reliability. In our evaluation framework, a branched schedule shifts the reliable tool group at hidden...

    arxiv.org/abs/2607.13396 · PDF

  21. 21

    EZSMT Version 3, Matured

    Yuliya Lierler

    cs.AI

    Constraint Answer Set Programming (CASP) is a hybrid reasoning paradigm that combines Answer Set Programming (ASP) with Constraint Processing and Satisfiability Modulo Theories (SMT), enabling powerful declarative encodings of complex combinatorial search problems. This paper presents the design and implementation of EZSMTV3, an extensible SMT-based CASP framework that advances the translational approach to CASP solving. Building upon the...

    arxiv.org/abs/2607.13344 · PDF

  22. 22

    Theory-Level Autoformalization: From Isolated Statements to Unified Formal Knowledge Bases

    Marcus J. Min, Mike He, Zhaoyu Li, Zixuan Yi, Sharad Malik, Aarti Gupta, Xujie Si, Osbert Bastani

    cs.AI · cs.CL · cs.LG · cs.PL

    Autoformalization translates informal natural language into formal, machine-verifiable languages. While most work focuses on individual statements, real formalization efforts are inherently theory-level: they require an entire web of axioms, definitions, and lemmas before target theorems can even be stated. In this position paper, we argue for theory-level autoformalization: formalizing complete theories, including all their...

    arxiv.org/abs/2607.13292 · PDF

  23. 23

    Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and Editable

    Ruhan Wang, Yucheng Shi, Zongxia Li, Zhongzhi Li, Yue Yu, Junyao Yang, Kishan Panaganti, Haitao Mi, Dongruo Zhou, Leoweiliang

    cs.AI · cs.SE

    The capability of a modern AI agent depends not only on its foundation model but also on its harness, which constructs prompts, manages state, invokes tools, and coordinates execution. As models, APIs, environments, and requirements evolve, the harness must be continually modified. Before such a change can be made, a developer or coding agent must identify all code locations that implement the target behavior. This is difficult because...

    arxiv.org/abs/2607.13285 · PDF

  24. 24

    Cost-Optimal Foundation Model Deployment Portfolio for Transportation Management

    Xi Cheng, Ke Liu, Siyuan Feng, Jane Lin, H. Oliver Gao

    cs.AI

    Foundation models, including large language models (LLMs) and vision-language models (VLMs), are increasingly used for transportation management center (TMC) tasks such as anomaly detection, incident reporting, and traveler information. Deploying multiple such models across TMC functions raises a portfolio question: which model should serve each function, in which deployment mode, and under what shared hardware budget? We formulate this as...

    arxiv.org/abs/2607.13239 · PDF

  25. 25

    AI-Native Insurance for Agentic AI: Pricing, Underwriting, and End-to-End Automation

    Quanyan Zhu

    cs.AI · cs.CR · cs.CY

    Agentic AI introduces new insurance challenges because autonomous AI systems can make decisions, invoke tools, modify external environments, and interact with third-party services. This paper develops an AI-native mathematical framework for underwriting, pricing, and contract design for agentic AI deployments. A deployment is represented by a risk state that captures autonomy level, operational authority, permission exposure, governance...

    arxiv.org/abs/2607.13230 · PDF

  26. 26

    Networked Intelligence: Active Shared Context Graphs for Human-AI Team Science

    Sutanay Choudhury, Jeffrey J. Czajka, Lummy M. O. Monteiro, Erin Bredeweg, Jason McDermott, Katherine Wolf, Alex...

    cs.AI · cs.CE · cs.HC

    Most AI-for-science systems focus on scaling a single reasoning process through better models, larger context windows, long-horizon agentic execution, or digital co-scientists working with one principal user. However, challenging scientific problems are rarely solved by one reasoner alone. They are solved by teams whose members bring different priors, experimental backgrounds, tacit knowledge, and domain-trained intuitions. The open problem...

    arxiv.org/abs/2607.13220 · PDF

  27. 27

    CayleyR: Solving the TopSpin puzzle via cycle intersection

    Yuri Baramykov

    cs.AI

    We present cayleyR, an R package for solving permutation puzzles by detecting cycle intersections in Cayley graphs. The core algorithm performs an iterative bidirectional search: from both the initial and target permutation states, random operation sequences generate cycles in the Cayley graph of the symmetric group Sn; their intersection yields a connecting path. When no direct intersection is found, a distance-guided bridge selection...

    arxiv.org/abs/2607.13219 · PDF

  28. 28

    Learning Safe Agent Behaviour from Human Preferences and Justifications via World Models

    Ilias Kazantzidis, Timothy J. Norman, Yali Du, Christopher T. Freeman

    cs.AI · cs.LG

    We address the problem of safely training an agent policy and deploying a good and safe policy, in settings where the environment dynamics are unknown and no suitable reward function is available. In the context of safety-critical environments, we consider traditional reinforcement learning impractical and resort to the resource of human input. We introduce DROPJ, a human-centred method for both safe training and deployment. We first learn a...

    arxiv.org/abs/2607.13172 · PDF

  29. 29

    Oracle Agent Memory as an Enterprise Memory Substrate for Long-Horizon AI Agents

    Richmond Alake, Cesare Bernardis, Paul Cayet, Luca Engel, Damien Hilloulin, Sungpack Hong, Allen Hosler, Nickolas...

    cs.AI · cs.DB

    Agent memory is a systems problem for long-horizon agents. Practical deployments require retention of task state across extended conversations, recovery of user-specific facts and preferences across sessions, and accumulation of procedural knowledge from prior outcomes. These requirements extend beyond document retrieval: a memory layer must determine which interactions become durable state, how that state is scoped, how it is retrieved under...

    arxiv.org/abs/2607.13157 · PDF

  30. 30

    Do AI Agents Know When a Task Is Simple? Toward Complexity-Aware Reasoning and Execution

    Junjie Yin, Xinyu Feng

    cs.AI · cs.CL · cs.SE · eess.SY

    Large language model (LLM) agents increasingly automate multi-step engineering and informatics workflows, yet they rarely ask how much effort a task actually requires. They often follow a maximum-context-first strategy--re-reading files and dependencies they have already seen--turning a one-line edit into a small code-base audit. We argue the missing capability is task-aware execution-scope estimation: judging a task's difficulty, the...

    arxiv.org/abs/2607.13034 · PDF

  31. 31

    Audio-Native Speech Recognition with a Frozen Discrete-Diffusion Language Model

    Harsha Vardhan Khurdula, Abhinav Kumar Singh, Yoeven D Khemlani, Vineet Agarwal

    cs.AI · cs.SD

    Automatic speech recognition is dominated by autoregressive decoders that emit one token at a time. We ask whether a discrete diffusion language model can transcribe speech instead, refining a whole transcript in parallel over a small number of denoising steps. We train an audio-native interface for DiffusionGemma, a 26B mixture-of-experts model that generates text by uniform, random-token discrete diffusion rather than the absorbing-mask...

    arxiv.org/abs/2607.13013 · PDF

  32. 32

    Dynamic Resource Allocation for Ensemble Determinization MCTS

    Jakub Kowalski, Adam Ciężkowski, Artur Krzyżyński, Mark H. M. Winands

    cs.AI

    Simulation-based algorithms are especially suited for high-uncertainty environments such as adversarial board games with significant elements of randomness and hidden information. In particular, several Monte Carlo Tree Search (MCTS) variants are commonly used in such domains. In this paper, we propose a series of enhancements for Ensemble Determinization MCTS, introducing two axes for dynamic resource allocation. First, Dynamic Number of...

    arxiv.org/abs/2607.13007 · PDF

This edition is part of The Daily Abstract — cs.AI archive. Subscribe to receive these in your inbox each morning, automatically translated to Spanish, with reply-to-PDF: arxivdaily.ignorelist.com.

Colophon Set in Georgia, with system sans for interface chrome and a monospaced stack for code and paper identifiers. Sole accent: amber #D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.