The paper introduces TEAM-Design, a rule that assigns two replay probabilities to each task—one for a human-only replay and one for an agent-only replay—based on how difficult it is to predict the missing baseline outcome and the cost of replay. It addresses the challenge of deciding whether to keep a human-AI workflow or replace it with a single actor when only one outcome can be observed after deployment. The authors prove that TEAM-Design solves the budgeted design problem and controls error rates, and demonstrate its effectiveness on clinical and coding benchmarks, noting it excels when one comparison is clearly harder than the other.
By Hamed Khosravi, Xiaoming Huo
PopResume is a population‑representative resume dataset designed for causal fairness auditing of large language model (LLM) and vision‑language model (VLM) resume screeners. It grounds fairness evaluation in real population statistics and preserves natural attribute relationships, enabling path‑specific effect (PSE) analysis that separates business‑necessity from redlining pathways. Using PopResume, the authors evaluated eight models on 60.8K resumes across five occupations and uncovered five discrimination patterns that aggregate metrics missed, demonstrating the value of causally‑grounded auditing.
By Sumin Yu, Juhyeon Park, Taesup Moon
RubricRefine is a training‑free pre‑execution refinement method that generates task‑specific rubrics from tool documentation, scores candidate code against explicit contract checks, and iteratively repairs failures before execution. It achieves an average score of 0.86 across seven models on M3ToolEval without any execution attempts, outperforming prior inference‑time baselines while incurring lower latency. The approach shows consistent performance on single‑step API‑Bank tasks and maintains an advantage in multi‑turn settings on AppWorld, with its effectiveness tied to the quality of the supplied documentation.
By Will LeVine, Brendan Evers, Sam Saltwick, Abhay Venkatesh
The paper investigates how different components of a graph retrieval‑augmented generation pipeline affect large language model performance on knowledge‑graph question answering. It examines four variables—whether the answer path is included, the syntax of triples, the order of triples, and the subgraph size—across six LLMs and two benchmarks. The study finds that including the answer path is crucial, while the grounding instruction dramatically reduces accuracy when no facts are provided, and that syntax, order, and subgraph size have negligible measurable impact at multi‑hop depth.
By Arquimedes Canedo
The paper introduces PUID, a Personalized Unobserved-Confounding-aware Interaction Deconfounder, designed to mitigate hidden confounding in recommender systems without relying on costly randomized controlled trials. PUID estimates user-item level sensitivity bounds using an entropy-based method that gauges the strength of hidden confounding from the mutual information between observed features and exposure status. An adversarial optimization strategy and a benchmark-guided variant (BPUID) further enhance robustness and predictive accuracy, and experiments on three real-world datasets show consistent outperformance over state-of-the-art baselines.
By Zongyu Li
The study evaluates how the choice of test boundary affects feature‑based hardware Trojan detection across Trust‑Hub families. Using a corpus of 49,124 gates from 16 netlists, the authors compare three test settings—pooled gates, a single netlist held out, and an entire host family held out—showing that performance drops markedly when a host family is excluded. The results demonstrate that sibling benchmark variants can inflate detection metrics, and the authors recommend reporting family‑aware holdouts alongside pooled scores.
By Hang Xiao, Chuhong Xu, Kainan Zhou, Gangzhen Qian, Lu Yi
The paper introduces AMA, a framework that uses multiple agents—Constructor, Retriever, Judge, and Refresher—to manage memory for large language model agents. AMA’s hierarchical memory design dynamically adjusts retrieval granularity to match task complexity, while the Judge and Refresher ensure relevance, consistency, and timely updates. Experiments on long-context benchmarks show AMA outperforms existing baselines and cuts token usage by about 80% compared to full-context approaches.
By Weiquan Huang, Zixuan Wang, Hehai Lin, Sudong Wang, Bo Xu, Qian Li, Beier Zhu, Linyi Yang, Chengwei Qin
APPSim-Bench is a new benchmark for mobile GUI agents that uses controllable simulated apps to balance realism and reproducibility. It includes 557 tasks across 17 popular Chinese and English apps, with a coding-agent-assisted and human-verified workflow that ensures deterministic evaluation. Evaluation of 19 agents shows that autonomous mobile execution is still far from perfect, with the best model completing only 50.27% of tasks and many failures in longer workflows and numerical reasoning.
By Jintian Feng, Long Chen, Xiao Yu, Jiayi Dai, Chenglong Liu, Haoru Wang, Zizhen Xue, Yuxuan Shi, Ziyang Wang, Yichen Gong
BuddyVQA is a new benchmark for companion‑style question answering on egocentric streaming video, comprising 21.6K questions tied to 6K highlight moments across 1,012 long first‑person videos. It emphasizes two often overlooked aspects of daily first‑person QA: ego‑deictic expressions and interactively chained questions, requiring models to resolve visual pronouns and infer user intent within a long‑form streaming context. The authors propose MyBuddy, a multimodal chain‑of‑thought QA assistant that uses a question filter and multi‑level memory to efficiently retrieve visual and QA information, achieving significant performance gains on BuddyVQA and generalizing to other streaming and common video QA benchmarks.
By Hangyu Qin, Junbin Xiao, Shenglang Zhang, Angela Yao
ReST‑RL introduces a unified Reinforced Self‑Training (ReST) policy‑value framework that enhances large language model (LLM) reasoning by combining an optimized ReST‑style GRPO algorithm with a value‑guided search (VM‑MCTS). The ReST‑GRPO component reshapes trajectory distributions to increase reward variance and expose policies to more informative partial states, improving training efficiency. VM‑MCTS trains a Value Model from self‑collected Monte‑Carlo Tree Search targets and uses it during inference to provide precise process signals and verification scores, boosting reasoning accuracy across coding benchmarks and out‑of‑domain math and science tasks.
By Sining Zhoubian, Dan Zhang, Jie Tang
arXiv:2609.09062v1 Announce Type: new
Abstract: We present a real-world case study of multi-task learning (MTL) for temporal process modeling from limited data with temporally sparse labels. Specific...
By Aseem Saxena, Paola Pes\'antez-Cabrera, Jonathan Magby, Markus Keller, Alan Fern
The paper introduces REASONS, a benchmark of 12,723 sentence-level citation instances across 12 arXiv subject categories, to evaluate scientific citation attribution under different evidence conditions. It proposes a dual-metric framework—Abstention Rate (AR) and Hallucination Rate (HR)—to balance reliability and responsiveness. Experiments with proprietary and open-source LLMs across various prompting and retrieval settings show that advanced Retrieval-Augmented Generation (RAG) reduces hallucinations but increases abstention, while adversarial metadata can push hallucination rates above 85%. Human evaluation confirms a high ratio of factual hallucinations to acceptable paraphrases, underscoring the need for systems that can appropriately abstain under uncertainty.
By Deepa Tilwani, Yash Saxena, Seyedali Mohammadi, Ankur Padia, Edward Raff, Amit Sheth, Srinivasan Parthasarathy, Manas Gaur
The study investigates how users interact with ChatGPT for code generation beyond simple function-level tasks, focusing on project-level benchmarks that involve multi-class dependencies. A user study with 36 participants examined prompting patterns, screen recordings, and chat logs to identify Human‑LLM Interaction (HLI) features that influence productivity. The results highlight three consistently supportive HLI features, five guidelines to boost productivity, and a taxonomy of 29 runtime and logic errors with mitigation strategies.
By Sangwon Hyun, Hyunjun Kim, Jinhyuk Jang, Hyojin Choi, M. Ali Babar
The paper critiques current evaluations of efficiency methods for large language model–based multi‑agent systems, arguing that reported gains are often inflated by method‑specific prompts and starting topologies. It introduces a controlled, MAS‑demanding diagnostic benchmark that standardizes the backbone model, agent registry, and runtime, and systematically varies topology, scale, depth, and tool use. The authors find that many claimed efficiency improvements are setup‑dependent, sometimes stemming from structural collapse or random pruning rather than genuine, robust gains.
By Jiamu Zhang, Lingxi Zhang, Pengjun Lu, Qiyue Zhang, Yu-Neng Chuang, Zhengchen Li, Shuai Xu, Vipin Chaudhary, Hanjie Chen
The paper introduces the Local Gradient Neural Operator (LGNO), a lightweight and interpretable neural operator designed for field temporal evolution prediction and source identification in mechanical problems. LGNO leverages nonlinear gradient discretization priors and multilayer perceptron convolutional layers to learn translation‑invariant local kernels resembling discrete stencils, with a zero‑consistent stencil factorization that separates coefficient learning from field reconstruction. Experiments on a range of PDE benchmarks—including linear, nonlinear, static, dynamic, low‑ and high‑dimensional cases—demonstrate that LGNO achieves comparable accuracy to global neural operators while using fewer parameters and maintaining rollout stability across diffusion, flow, and quantum phenomena.
By Baiming Zhang, Jinsong Tang, Ying Xu, Lihua Chen, Shiying Xiong
Shortcut learning occurs when classifiers rely on spurious correlations instead of true diagnostic features. The authors argue that this is essentially a calibration issue: unconstrained training aligns shortcut groups with their prevalence, causing over‑confidence in one group and under‑confidence in another. They introduce two encoder‑agnostic methods—a regularizer and a post‑hoc recalibration—to equalize prevalence across shortcut groups, achieving significant performance gains on chest‑drain‑pneumothorax benchmarks with both fine‑tuned CNNs and frozen foundation‑model backbones.
By Mohamed Amine Kina, Eike Petersen
The paper introduces RouteOPD, a new on‑policy distillation method that treats teacher‑student disagreement as a probability transport problem. By decomposing disagreement into excess sources and deficit destinations, RouteOPD pairs them explicitly and optimizes pairwise log‑odds toward targets derived from a bounded teacher potential. Experiments on four teacher‑student configurations and four mathematical‑reasoning benchmarks show that RouteOPD consistently outperforms sampled reverse‑KL OPD, achieving higher routing fidelity and lower background leakage.
By Tianle Xia, Lingxiang Hu, Yiding Sun, Linfang Shang, Ming Xu, Lan Xu, Ning Zheng, Wei Xu, Jie Jiang
ProcArena is a new benchmark for evaluating large language models on natural‑language to PL/SQL translation tasks. It contains 3,998 executable tasks across 157 databases, covering nine development subscenarios in PostgreSQL and Oracle, and supports both direct generation and interactive multi‑turn scenarios. Experiments on seven models show that even the best performers achieve only about 62% accuracy in direct mode and 58% in interactive mode, highlighting the difficulty of realistic NL‑to‑PL/SQL development.
By Hang Zhang, Chaokun Wang, Yuzhi Pan, Ziyao Zhong, Shuo Cao, Yue Xue, Zeyu Huang, Xingwei Zhou, Fang Niu, Bofan Xie, Guanchen Ge, Leqi Zheng, Ziyang Liu, Xiannian Cao, Pengcheng Ge
arXiv:2609. 07344v1 Announce Type: cross Abstract: Large language model (LLM) based agents are increasingly applied to cybersecurity tasks such as vulnerability discovery and automated penetration testing.
By Weizhe Wang, Yitong Zhang, Yao Zhang, Xiaoqiang Di, Zhigang Li, Bin Wu, Guangquan Xu
FedGenSC introduces a federated generative semantic communication system that uses a global generator with local discriminators, a semantic prototype bank, and SNR‑conditioned generation to address instability, semantic drift, and channel‑agnostic issues in GAN‑based federated learning. Experiments on the Europarl dataset over Rayleigh fading channels show that FedGenSC outperforms the FedDeepSC baseline under non‑IID data, achieving up to a 58.2% relative improvement in BLEU‑1 at 18 dB. Ablation studies confirm that each component independently contributes to the overall performance gains.
By Rita Abou Fares, Razan Al Kakoun, Maher Nouiehed, Hadi Sarieddeen