DA‑WAM is a framework that integrates predictive representation learning, action‑conditioned future modeling, and trajectory scoring into a single decision‑making objective for autonomous driving. It uses an online encoder with a stable momentum target to keep future representations aligned with the driving task, generating a distinct future latent for each trajectory candidate. A future‑latent‑conditioned scorer evaluates these latents, with expert‑matched trajectories supervised by observed futures and safety‑critical hard negatives providing additional guidance, achieving state‑of‑the‑art results on NAVSIM‑v1 and NAVSIM‑v2.
By Ruiguo Zhong, Benshan Ma, Xiaolong Chen, Lang Zhang, Mingyue Feng, Yaonong Wang, Pei Liu, Jun Ma
Flama is an open‑source Python framework that unifies the development and deployment of production‑ready web APIs, machine‑learning services, and large‑language‑model (LLM) applications. Built on ASGI, it offers an async‑first, type‑driven programming model with seven subsystems—including dependency injection, a pluggable schema layer, automatic CRUD generation, a portable binary model format, a multi‑backend LLM server, a Rust‑accelerated core, and a Model Context Protocol module. The framework also provides built‑in JWT authentication, pagination, background tasks, WebSocket and streaming support, OpenAPI generation, and a CLI for running, packaging, and inspecting models.
By Jos\'e A. Perdiguero L\'opez, Miguel A. Dur\'an-Olivencia
The paper investigates how large language models can extract contextualized data from scientific literature. It presents four workflows: expert‑written prompts, self‑generated prompts, autonomous literature discovery, and dataset creation from guidelines. While models perform well with prompts, they struggle with context, hallucinate references, and still need human oversight for final validation.
By Valentin Romanov, Monique Bax, Steven Niederer
The article reviews the emergence of Agentic AI, covering its evolution, theoretical foundations, working principles, and architectural aspects. It surveys recent scholarly contributions across various domains, highlighting real‑world applications, current research findings, and existing challenges. The review also proposes a framework for stakeholder adoption and outlines future research directions to guide researchers and practitioners.
By AKM Bahalul Haque, Al Amin Islam Ridoy, Mohammad Rayhan, Ivan Porres
The paper introduces a multi‑agent platform built on CrewAI for conversational business intelligence. Five specialized agents process natural language queries, retrieve and analyze data, generate visualizations via the Model Context Protocol, and deliver actionable insights. The system includes a defense‑in‑depth security architecture, a query parameterization mechanism, and achieves 95.3% functional accuracy with a 24‑second mean latency, outperforming a single‑agent baseline by 22.6 percentage points in accuracy and 20.2% in quality.
By Manoj N M, Vijayakrishna S, Manjunath Srinivas, Rohit Pahan
Sanyu Studio is a multi‑agent dialogue system that treats 321 Sanyu oil paintings as agents equipped with fact, interpretation, organization, and memory‑filtering mechanisms. The paper reports on a seven‑day workshop with eight art‑university participants, showing that user prompts, evidence organization, and cognitive tendencies produced divergent yet coherent digital narratives of Sanyu. The study suggests that, when historical evidence is limited, AI can amplify human agency and provide public audiences with an interactive entry point into art‑historical interpretation.
By Zhaoxi Wei, Hongye Yang, Shuyuan Tian
FACET is a framework for synthesizing terminal tasks that preserves source intent and ensures cross‑artifact consistency. It reconstructs agent skills into coherent scenarios, repairs the execution environment, and uses the resulting container state as shared grounding for the instruction, solution, and verifier. By validating and repairing artifacts through execution, FACET produces complex tasks with dense executable checks and data‑efficient supervision, improving performance on Terminal‑Bench 2.1.
By Kou Shi, Zun Wang, Qisheng Su, Shiting Huang, Ziao Zhang, Zhen Fang, Qingnan Ren, Jin Liu, Yu Zeng, Yiming Zhao, Lin Chen, Zehui Chen, Feng Zhao
The paper introduces the Orienteering Problem with Uncertain Time‑Varying Rewards (OP‑UTVR), a new variant of the classic orienteering problem that allows agents to estimate and forecast reward dynamics from observations. Three planners with different planning horizons and online adaptivity are proposed, and theoretical performance bounds under reward stochasticity are derived. A mobile service robot benchmark is presented, and experiments show trade‑offs between planning horizon and adaptivity, highlighting the benefits of long‑horizon planning with online adaptation.
By Masafumi Endo, Kohei Honda, Yuu Jinnai, Ryo Yonetani
The paper proposes a theory for judging post-hoc debates in AI, focusing on properties like reproducibility, robustness, groundedness, and explainability. It evaluates two debate‑judgement methods—LLM judges and formal computational argumentation semantics—finding similar accuracy but noting that argumentation semantics offers stronger formal guarantees. The study suggests that argumentation semantics is a preferable framework for principled debate judges in AI systems.
By Xiang Yin, Adam Dejl, Antonio Rago, Lihu Chen, Francesca Toni
The paper introduces EvoResearcher, a training‑free, inference‑time protocol that enables a frozen large language model to perform cost‑bounded self‑reflection and early stopping. By iterating through generate → self‑critique → revise steps until a maximum depth or a CONFIRMED sentinel is reached, the model can self‑verify its answers within a strict compute budget. The protocol incorporates four self‑reflective meta‑reward components—correctness, efficiency, reflection depth, and tool‑call diversity—implemented as prompt‑level mechanisms, and is validated on Big‑Bench Hard, GSM8K, and MATH benchmarks, achieving comparable accuracy while terminating 82‑88% of items early with only about 2.1 generations per question.
By Wei Yu, Suxing Liu, Minjie Yu, Jiahao Wang, Zhijian Zheng, Haocheng Deng, Bing Li
DeepTCM1.0 is a multi‑expert AI agent built on the DeepSeek V3.2 large language model, designed to integrate classical traditional Chinese medicine (TCM) theory with modern life sciences for mechanistic analysis of TCM compound formulas. The framework employs a three‑tier collaborative architecture and a three‑round iterative quality‑control workflow, simulating 11 interdisciplinary intelligent agents. It was validated on the Guizhi Decoction, with performance evaluated through double‑blind five‑dimensional scoring, ICC reliability testing, Mann‑Whitney U tests, and effect size analysis across 100 independent assessments by four large language models.
By Wenxin Duan, Hanwei Wang, Zhongying Peng, Zhonghua Lu, Jiayi An, Fan Song, Yong Liang
The paper introduces PAFIR, a Personalized and Adaptive Feature selection framework that treats feature selection as a reinforcement learning problem over longitudinal multimodal health data. PAFIR jointly models structural dependencies among assessment variables and temporal dynamics in wearable-derived physical activity data, learning adaptive selection policies across repeated study visits using reward signals from sparse fall incidence outcomes. Applied to the Physio Feedback Exercise Program (PEER) trial, PAFIR outperforms state‑of‑the‑art baselines by capturing longitudinal and structural patterns of feature relevance, enabling dynamic, subject‑specific feature selection for more timely fall prevention strategies.
By Chang Liu, Ladda Thiamwong, Yanjie Fu, Rui Xie
The paper investigates how the CutMix data augmentation technique affects reliability and robustness in semantic segmentation. It evaluates two architectures—CNN-based DeepLabV3+ and transformer-based SegFormer—on both in-domain and out-of-domain data. Results show that CutMix has a minor effect on segmentation accuracy but consistently improves reliability, especially under distribution shifts, by enhancing calibration and uncertainty quality.
By Steven Landgraf, Markus Ulrich
TestifAI is a deep learning testing framework that estimates model robustness against combinations of semantic input perturbations such as blur, brightness, and zoom. It allows users to define operational conditions as structured spaces with discrete severity levels and query robustness for any combination. By employing partial model tomography, TestifAI reconstructs higher‑order perturbation effects from low‑order tests, achieving less than 7% estimation error while reducing inference counts by 60‑80% across five image and language classification tasks.
By Arooj Arif, Tobias Hartung, Elena Botoeva, Alexandros Koliousis
The paper proposes a verifiable abstention framework for AI-driven leak localization in water distribution networks, where a physics-based executor agent tests leak hypotheses against a digital twin and an independent supervisor with an LLM auditor certifies actions or abstains. In noisy field conditions, the system achieves 96% decision precision on acted events, correctly identifies all leaks in a benchmark, and demonstrates practical deployment with a 194-event audit record. The approach offers a defensible, accountable method for autonomous water‑infrastructure operation.
By Tianwei Mu, Yue Wang, Mingzhe Yuan, Manhong Huang, Wenhong Wang, Xuerui Yin, Qing Luo, Min Xiao, Hui Yang, Jun Li, Dan Xue
LEDGER is a tracing and review system for large language model agents that constructs layered trace graphs from observed sessions. It groups raw trace records into Evidence Nodes and Workflow Nodes, anchors artifacts as evidence, and adds typed semantic edges linking claims to supporting actions, artifacts, and checks. The resulting traces reveal workflow decisions, artifact lineage, repair steps, validation coverage, and claim‑support paths for evidence‑centered audit.
By Daehong Kim, Haichao Miao, Shusen Liu
SeisEvo is a method that uses a large language model (LLM) and multi‑agent search to evolve seismic data reconstruction algorithms rather than optimize a single result. Starting from a classical algorithm, the agents modify only user‑opened components, rejecting candidates that violate physical constraints and scoring the rest by execution. The resulting white‑box algorithms—such as a residual‑gated, phase‑aligned dip‑consistency projection for interpolation and a reliability‑grouped singular‑value shrinkage for simultaneous interpolation and denoising—outperform classic methods by several decibels and generalize to unseen data.
By Yingjie Xu, Siwei Yu, Jianwei Ma
The paper proposes a new approach called reversible forgetting for continual learning in enterprise AI agents. It introduces a framework with three memory states—active, dormant, and retired—and a reactivation transition that can restore dormant knowledge when it becomes relevant again. The authors implement this as a Hysteretic Reversible Memory Controller that uses relevance evidence, asymmetric thresholds, shadow mode testing, and policy gating to reduce the influence of obsolete information without permanently erasing it.
By Nilutpaul Sarker Yash, Tirtho Roy, Ushashi Bhattacharjee
ChainWorld is a new benchmark that composes atomic OSWorld tasks into long‑horizon desktop workloads, creating 347 chains of length two to four that compare two renderings of the same task sequence. In single‑turn evaluation all tasks are presented together in one prompt, while in multi‑turn evaluation tasks are revealed one at a time. Across four current computer‑use agents, maximum chain completion is 31 %, with multi‑turn evaluation improving completion for three models but both protocols remaining challenging and exposing different failure profiles.
By Vincent Siu, Manasi Sharma, Dawn Song, Daniel Yue Zhang, Chenguang Wang, Ying Liu
EgoMemReason is a new benchmark for week‑long egocentric video understanding that focuses on memory‑driven reasoning rather than simple perception tasks. It tests three memory types—entity, event, and behavior—across 500 questions, each requiring evidence from an average of 5.1 video segments and 25.9 hours of backtracking. Evaluation of 17 models shows that even the best achieves only 39.6% accuracy, highlighting the difficulty of long‑horizon memory in multimodal systems.
By Ziyang Wang, Yue Zhang, Shoubin Yu, Ce Zhang, Zengqi Zhao, Jaehong Yoon, Hyunji Lee, Gedas Bertasius, Mohit Bansal