The paper introduces FARSIGHT, a framework for evaluating the robustness and security of financial trading agents powered by large language models. It assesses agents on their resilience to market turbulence, such as flash crashes, and their vulnerability to three types of attacks: on information sources, on the agents themselves, and on agents acting as attackers. Applying FARSIGHT to 15 academic trading schemes reveals that 80% fail at least one robustness test and all exhibit security weaknesses, highlighting the risk of market-wide crashes from both accidental misjudgments and deliberate attacks.
By Mengxiao Wang, Nitesh Saxena
The paper introduces an end‑to‑end framework that automatically builds a business semantic layer from raw application logs. It uses a two‑stage abstraction: first, large language models identify high‑level business features with industry knowledge, and second, a structured pipeline refines data, retrieves relevant information, filters, clusters semantically, and assigns canonical names. Evaluation on production‑scale telemetry shows significant gains in semantic quality, noise reduction, and maintenance effort, with a 0.87 Cohen’s kappa in an LLM‑as‑Judge assessment.
By Yuanzhe Jia, Ali Anaissi
DataCanvas-EDU is an agentic framework that lets instructors guide the creation of synthetic datasets for business analytics courses. Instructors set teaching goals and desired patterns via conversation, and an AI agent writes generation code, verifies the data, and produces assignments, reference solutions, and rubrics. The process is organized into four phases—Plan, Create, Verify/Test Analysis, and Evaluate—to streamline case preparation and enable students to explore new patterns with AI.
By Bang An, Maria Hamdani, Joseph Fox
The paper critiques the common practice of evaluating large‑language‑model (LLM) evolutionary search methods using a single seed and fixed iteration budget, arguing that this approach is insufficient. By testing three search strategies across five optimization tasks and varying both the number of seeds (width) and iterations (depth), the authors find that optimal budget allocation depends on the strategy, task, and total budget, and that strategy rankings shift with different budgets. They propose a measurement protocol that maps the seeds‑by‑iterations frontier and offers practical guidance for researchers.
By Tal Oved, Roi Pony, Oshri Naparstek, Udi Barzelay
Zarya is a hybrid language model that jointly trains an autoregressive objective and a masked-diffusion objective within a single architecture. It structures training data into variable-size slots and uses a curriculum that gradually increases slot granularity, allowing a smooth transition from fine-grained AR learning to coarse-grained diffusion learning. At inference, Zarya offers two decoding modes—MDM sampling with first-hitting denoising and slotted speculative decoding that interleaves diffusion-based selection with autoregressive infilling—while fully decoupling training and inference regimes and supporting extensive configurability.
By Leonid Sinev, Ilya Koziev, Vladislav Leshchuk
DeliveryGym is a 3D reinforcement learning environment that simulates continuous courier shifts, integrating multimodal tool interaction, persistent world dynamics, and trajectory‑based rewards derived from simulator events. It allows agents to learn how their decisions affect time, energy, and money across an entire shift, and it adapts future training shifts to the policy’s weaknesses while keeping evaluation fixed. Experiments on six models and 13 city maps show a significant gap between task execution and optimal sequencing, with RL improving Qwen3‑VL‑4B’s net income by 54.3% and adaptive training boosting test income by 16.5% over uniform sampling.
By Haoqiang Kang, Yiming Zhang, Yiyang Guo, Chuying Li, Jianzhi Shen, Tianruo Rose Xu, Xiaokang Ye, Lianhui Qin
MATCH is a closed‑loop framework for model‑aware tool learning that combines curriculum scheduling with hierarchically gated rewards. It introduces Model‑Aware Curriculum Learning (MACL), which dynamically adjusts sample difficulty based on reward signals, and Hierarchical Tool‑call Gated Reward (HTGR), which allocates credit at the tool name, argument key, and argument value levels only when prerequisites are met. Experiments on API‑Bank and BFCL V3 show MATCH achieving 72.19% and 62.87% overall accuracy, outperforming both supervised and RL‑based baselines across multiple backbone models.
By Shihao Liu, Hao Yin, Lijun Liu, Zhengzong Chen, Yuanyuan Zhao, Fei Huang
AdaRepair-Mem introduces an adaptive experience retrieval framework for repository-level program repair, addressing three key limitations in existing memory-augmented methods: imbalanced episodic memory, non-monotonic success with increased memory, and phase misalignment of memory types. The framework employs coverage-aware retrieval, quality-aware selection, and stage-aware routing to better match repair contexts with relevant memories. Evaluations on SWE-Bench-Lite and SWE-Bench-Verified show improved performance on under-covered repositories, reduced noisy retrieval, and enhanced patch refinement support.
By Z. C. Luo, J. C. Guo, W. J. He, S. Y. Wang, J. C. Yu, F. M. Zhao, Y. Chen, T. Cao, L. Q. Liu, N. Zheng, W. Xu, J. Jiang, Z. M. Zhao
The study investigates how to prevent skill loss from relying on AI assistants by testing two interventions. In a preregistered online experiment with 704 participants, metacognitive feedback—explicitly highlighting the consequences of offloading—significantly lowered the likelihood of users offloading answers to an LLM and improved their performance on a subsequent unaided test. No effect was found for an effort-based reward incentive. The findings suggest that metacognitive feedback is an effective design strategy to reduce cognitive offloading to AI.
By Sebastian Maier, Kai Schwabe, Manuel Schneider, Stefan Feuerriegel
The paper introduces DB‑SUiT, a surface‑based diffusion bridge that translates cortical MRI to PET images directly on the cortical manifold. It employs a conditional spherical U‑shaped vision Transformer to capture multi‑scale surface features and long‑range dependencies while incorporating demographic and subcortical information. Evaluations on two dementia datasets show that the synthesized PET surfaces outperform MRI and PET volumes in automated classification and achieve high diagnostic accuracy in a blinded reader study.
By Yitong Li, Alexandra Samoylova, Fabian Bongratz, Timo Grimmer, Dennis M. Hedderich, Igor Yakushev, Christian Wachinger
The paper presents the first music‑specific, layer‑wise empirical study of hallucination in audio‑language models, framing it as a hierarchical perceptual grounding failure across five layers: sound events, temporal properties, tonal attributes, style, and emotion. It introduces MuseDiag, a diagnostic framework that evaluates nine models and finds universal vocal misperception, significant tonal perception differences, and identifies Audio‑Flamingo‑3 as the most stable model. The study also proposes two training‑free mitigation methods, ADD‑M and TPA, which reduce hallucination in probing but show variable effectiveness in free‑form generation, highlighting the need for multi‑paradigm evaluation.
By Yu Liu, Jiahui Liu, Zhilin Liu, Cong Cao, Fangfang Yuan, Yuling Yang, Pin Xu, Yanbing Liu
The paper investigates whether training classical machine learning models remains worthwhile when large language models (LLMs) can label tabular data without training. By defining a labeled‑data crossover point (N*) where a trained classical model surpasses a frozen LLM’s flat error, the authors analyze 126 student evaluations of GPT models across 18 datasets and compare them to power‑law learning curves of six classical model families. Results show that in 86% of cases a classical model outperforms the LLM with no more labeled data than already available, and the crossover occurs at a median of about 6% of the training set, suggesting that collecting a few hundred labels and training a gradient‑boosted model is typically advantageous.
By Kaihua Ding
The paper introduces an automated pipeline that converts non‑critical driving scenes into safety‑critical scenarios by integrating computer vision, Large Language Models (LLMs), and Augmented Reality (AR). It detects and tracks road users, extracts safety features such as distance, velocity, motion direction, and Time‑to‑Collision (TTC), and evaluates scene criticality. Safe scenes are then modified by an LLM, which generates realistic collision‑inducing objects and behaviors that are overlaid onto the original scene using AR, achieving 97.52% safety classification accuracy on the nuScenes dataset and producing realistic scenarios like pedestrian crossings, rear overtaking vehicles, and sudden‑stop events.
By Noura Fady, Farah Khaled, Catherine M. Elias
STR-Agent is an LLM-driven framework designed for QoS-aware routing in Low Earth Orbit satellite networks. It integrates intent perception, tool-based execution, experience accumulation, and reflection-based policy adaptation to translate natural-language service requests into adaptive routing decisions. In simulations on a Walker-Delta constellation, STR-Agent reduces end-to-end delay by up to 60% compared with DQ-Dijkstra and improves intent-understanding accuracy from 45.4% to 92.45% after fine-tuning, with the Reflection Module providing additional delay reductions.
By Bowen Lu, Mugen Peng, Yaohua Sun, Hongyu Wang, Kerui Guo, Wenjia Xu
The paper introduces SPAR, a closed‑loop simulation platform that couples real‑time AUV control software with a higher‑level orchestration layer for fault injection, prompting, and evaluation of large language models (LLMs) in diagnosing and recovering from anomalies. SPAR enables ensemble testing of LLMs, comparing a frontier model with three locally deployable LLMs on a mass‑shift fault scenario across 480 trials, revealing that model choice significantly affects diagnostic accuracy. The study demonstrates that while the frontier model consistently ranks the correct fault mechanism among its top hypotheses, local models succeed mainly when they follow the full diagnostic procedure, and overall diagnosis and operational decisions appear decoupled in this dataset.
By Khalid Halba, Kylie Cooper, James G. Bellingham
FAMOS is a feed‑forward model that predicts movable‑part segmentation and joint parameters from a sparse, unordered set of partial point clouds. It jointly reasons over multiple observations using a Multi‑state Articulation Transformer that alternates state‑wise and global attention, and introduces an observed articulation span objective to supervise motion ranges across inputs. A procedural data generator supplies self‑annotated assets for training, and experiments on PartNet‑Mobility, ACD, and ArtiCraft‑10K show consistent improvements over existing feed‑forward and optimization‑based baselines.
By Kevin Qu, Tao Sun, Massimiliano Viola, Liyuan Zhu, Zhizhuo Zhou, Sayan Deb Sarkar, Konrad Schindler, Iro Armeni
TripScore is a benchmark and evaluation framework for large language models (LLMs) in travel planning, built from real user logs and calibrated with 1,468 pairwise judgments from 203 travel experts. It uses a hierarchical feasibility gate for format and commonsense checks, and a unified point-wise reward that combines soft quality and preference fulfillment. Experiments show that reinforcement learning fine‑tuning, such as GRPO, consistently outperforms other methods when evaluated with TripScore.
By Yincen Qu, Huan Xiao, Feng Li, Gregory Li, Hui Zhou, Xiangying Dai, Xiaoru Dai, Xuan Huang
The paper introduces SafeHarness, an obstacle‑aware framework that improves the safety of coding agents for robot manipulation. By decomposing tasks into route planning and contact execution, the harness enables the agent to prioritize collision avoidance, achieving 71.9% task success and 87.5% collision avoidance—significantly better than prior methods. The study demonstrates that safety constraints can be effectively integrated into language‑model‑driven robot controllers.
By Bingxin Xu, Yuzhang Shang, Zhen Dong, Emilio Ferrara
The paper introduces a method for selecting the best vision‑language model for a downstream task by analyzing the internal dynamics of the visual encoder. It represents each task with layer‑wise conductance and uses an entropy‑regularized alignment to derive a target‑conditioned block importance distribution. The proposed Directional Conductance Divergence (DCD) metric captures asymmetric transferability, enabling accurate prediction of model rankings without direct inference, and achieves a 14.7% NDCG@5 improvement over SWAB on 48 VLMs across 21 datasets.
By Wei Yang, Hong Xie, Tao Tan, Xin Li, Defu Lian, Enhong Chen
The paper argues that the architecture of multi‑agent large language model (LLM) frameworks, rather than just the intelligence of the underlying models, largely determines system performance. It introduces a taxonomy of architectural dimensions—such as orchestration, memory, planning interfaces, specialization, and communication topology—and presents MAFBench, a unified evaluation suite. An empirical study across nine frameworks, keeping the LLM constant, reveals six design principles and shows that choices like orchestration and communication topology can dramatically affect latency, accuracy, and coordination success.
By Abdelghny Orogat, Ana Rostam, Essam Mansour