The paper introduces Drift‑Aware Sparse Routing (DRS), a method for routing requests in a multi‑model language service while respecting compute, latency, memory, or cost budgets. DRS estimates reward and resource use from a rolling audit window, routes using pessimistic reward and optimistic cost estimates, updates resource shadow prices online, and applies a hard meter before commitment. The authors provide theoretical regret bounds that separate control from statistics, showing how the method adapts to non‑stationary prompt distributions and model changes.
By Cheung Hao Lee, Patrick Wong
The paper introduces EvoPrompt, a framework for adapting vision‑language models to new tasks with limited data while preventing catastrophic forgetting. EvoPrompt uses a Modality‑Shared Prompt Projector to create hierarchical prompts and an evolutionary training strategy that separates low‑rank updates into directional and magnitude components, preserving learned semantic directions. Experiments show that EvoPrompt achieves state‑of‑the‑art few‑shot performance while maintaining the original zero‑shot capabilities of the pre‑trained models.
By Enming Zhang, Jiayang Li, Yanlong Wang, Yanru Wu, Zhenyu Liu, Yang Li
The paper introduces SAM3-LoRA, a parameter‑efficient adaptation of the SAM3 segmentation foundation model using Low‑Rank Adaptation for multi‑class structural defect segmentation. It presents a supervision method that trains the model directly from COCO‑style instance segmentation by using category names as prompts, eliminating the need for prompt templates or learned embeddings. The authors also identify a failure mode where the model responds to any prompt due to positive‑prompt‑only annotations and resolve it with exhaustive hard‑negative prompting, achieving significant improvements in pixel IoU and instance recall on both a tunnel lining dataset and the Structural Defects Dataset.
By P. Malaisree, S. Youwai, S. Janrungautai, D. Amorndechaphon, P. Rojanavasu, W. Songkitti
Athena is a graph-based system that identifies affected libraries for software vulnerabilities by modeling vulnerability databases as a knowledge graph and applying knowledge graph completion. It integrates CVEs, libraries, CWE types, CPE products, and software ecosystems, then predicts missing affected libraries using link prediction and refines results with a fine‑tuned LLM that incorporates graph embeddings. Experiments on the VulLib dataset show Athena outperforms four state‑of‑the‑art baselines, achieving a 32% higher average F1 score and demonstrating that a smaller KGC backbone can surpass larger LLM‑only approaches.
By Phong Trinh Duy, Trang Dang Yen, Hung Nguyen-Huu, Bach Le, Quyet-Thang Huynh, Dieu Hoang Vu, David Lo, Thanh Le-Cong
The paper introduces Inverted Contrastive Learning for Unsupervised Feature Selection (ICLFS), a method that treats each feature as a sample by inverting the data matrix and applies a contrastive learning framework to learn consistent representations across masked positive views and a shuffled negative view. Feature saliency is derived from the magnitude of projector‑space embeddings, and a Laplacian‑Gated Ranking Correction step refines the ranking by reducing local redundancy. Experiments on 12 benchmark datasets show that ICLFS achieves the best clustering accuracy on 10 datasets compared to both classical and neural baselines, demonstrating the effectiveness of feature‑wise contrastive consistency for unsupervised feature selection.
By Utsab Ghosh, Roshni Chakraborty
APEX-EM is a non‑parametric experience memory that stores full procedural‑episodic traces in a typed Procedural Knowledge Graph and retrieves them via semantic search, structural‑signature matching, and graph traversal. It uses a Plan‑Retrieve‑Generate‑Iterate‑Ingest workflow to produce, quality‑gate, and commit experiences, indexing both successes and failures so the agent learns what to reuse and what to avoid. Evaluations on five benchmarks with a shared GPT‑4o backbone show significant performance gains, such as +7.6 pp on BigCodeBench transfer and +1.4 pp on Lifelong Agent Bench, demonstrating that the memory adds to model capability rather than replacing it.
By Pratyay Banerjee, Masud Moshtaghi, Ankit Chadha
EvoFlint is an evolutionary atlas that maps multi‑turn vulnerabilities in large language models by treating attack discovery as a search problem rather than a generation task. It uses evolutionary quality‑diversity search to evolve phased conversation plans, employing Pareto fitness for success rate and severity, novelty search for diversity, and a generation‑level memory to incorporate model insights. The resulting risk‑indexed archive, tested on HarmBench, shows high attack success rates across models such as Claude Sonnet, GPT‑5, and Qwen3, revealing which harm categories each model’s safety training covers or misses.
By Feitong Qiao, Liren Peng, Shiming Ren, Aishwarya Jadhav, Arghavan Bahadorinejad, Marinette Chen, Muhan Zhang, Abdulaziz Suria, Gennevi Lu, Anish Das Sarma
The paper introduces QXymb, a framework for building observational declarative twins of quantum circuits, and presents its first complete order‑0 specialization, QILP‑0. QILP‑0 transforms observed circuit behavior into a finite multi‑valued propositional logic program by incrementally traversing a declared family of quantum observables, quantifying progress via reference‑relative coverage, and preserving observational semantics through deterministic mapping back to original observable columns. Validation on Bars & Stripes and MNIST quantum machine learning settings shows that the induced QILP‑0 theory achieves perfect, conflict‑free reconstruction of the discrete relations, with logical exactness separated from numerical and discretization uncertainties.
whyItMatters":"The work demonstrates a method to construct exact observational declarative twins of quantum circuits, enabling precise logical reconstruction of quantum behavior independent of numerical uncertainties."
By Marina de la Cruz Echeand\'ia, C\'esar Luis Alonso, Tony Ribeiro, Alfonso Ortega de la Puente
VerTox is a framework that turns corpus poisoning of neural ranking models into a verifiable reward‑guided reinforcement learning problem. By fine‑tuning compact large language models with reward shaping that couples ranking distortion and factual corruption, VerTox generates fluent, low‑perplexity adversarial documents that frequently outrank target items across multiple ranking architectures, including a commercial embedding model. Experiments show near‑perfect attack success and significant degradation of downstream retrieval‑augmented generation performance.
By Zhiqi Huang, Vivek Datla, Zhichao Xu, Puxuan Yu, Vivek Srikumar, Alfy Samuel
The paper introduces a retrieval‑augmented generation framework that combines fine‑tuned open‑weight language models with a hybrid dense‑lexical retrieval pipeline to synthesize corrosion knowledge. Applied to magnesium alloy corrosion, the system achieves significant improvements in token‑level accuracy and high faithfulness and context recall, while an additional Reason Map framework constructs evidence graphs to detect causal errors that standard factuality metrics miss. The modular design is positioned as a general blueprint for trustworthy AI‑assisted knowledge synthesis across engineering domains.
By Bharath M N, R K Singh Raman, Alankar Alankar
arXiv:2609.01564v1 Announce Type: cross
Abstract: Large language models (LLMs) struggle to classify text into taxonomies with many semantically similar labels, as the distinctions are domain-specific...
By Manish Gupta, Chaitanya Giri, Jayasimha Talur
The paper introduces ExpArt-KG, a knowledge graph tailored to the artwork domain, and a retrieval‑augmented generation framework that alternates between generating answers and retrieving relevant facts from the graph. By using a correctness judgment to guide the search, the method efficiently gathers the necessary factual information, improving the detail of image explanations while reducing external knowledge retrieval costs. Experimental results demonstrate that the approach maintains generation quality comparable to fixed‑iteration methods.
By Yuta Kato, Shintaro Ozaki, Kazuki Hayashi, Yusuke Sakai, Hidetaka Kamigaito, Katsuhiko Hayashi, Taro Watanabe
The paper presents an end‑to‑end pipeline for automatically constructing a tree‑structured knowledge graph (KG) from Vietnamese high school history textbooks and evaluating retrieval strategies that exploit the KG’s hierarchical structure. The KG construction uses a three‑phase hybrid relation extraction process, including intra‑batch deduplication, approximate cross‑batch search, and LLM extraction with a centroid filter and dual‑LLM validator, resulting in 750 nodes and 4,341 semantic edges across 41 ontology types. Retrieval evaluation compares three graph traversal strategies—Top‑Down, Horizontal, and Bottom‑Up—on a benchmark of 1,210 Vietnamese queries, finding that the Top‑Down strategy with structural information outperforms a vector baseline by 4.7 percentage points in NDCG@10.
By Ket Doan Nguyen, Minh N. H. Nguyen
arXiv:2607.13345v2 Announce Type: replace
Abstract: We present a frame-independent audio-text system for the 3rd Ambivalence/Hesitancy Video Recognition Challenge at the 11th Affective & Behavior Ana...
By Luiz F. B. F. Martins, Rodrigo W. Pisaia, Matheus M. Girardi, Isabella V. Berkembrock, Jo\~ao A. Almeida, Andre G. Hochuli, Rayson Laroca, Alceu S. Britto Jr
arXiv:2604.08500v2 Announce Type: replace
Abstract: We tackle the problem of sparse novel view synthesis (NVS) using video diffusion models; given $K$ ($\approx 5$) multi-view images of a scene and t...
By Qi Wu, Khiem Vuong, Minsik Jeon, Srinivasa Narasimhan, Deva Ramanan
The paper introduces the Mass‑Conserving Perceptron (MCP), a physics‑aware AI framework that enforces conservation laws while learning hydrological process relationships from data. By progressively adding physically meaningful components—such as bounded soil storage, state‑dependent conductivity, variable porosity, infiltration capacity, surface ponding, vertical drainage, and nonlinear water‑table dynamics—to a single MCP storage unit, the authors demonstrate that predictive skill for daily streamflow improves across 15 U.S. catchments. The study finds that the impact of each process representation varies with hydroclimate, with vertical drainage boosting performance in arid and snow‑dominated basins but hindering it in rainfall‑dominated ones, while surface ponding has minimal effect; the best MCP configurations rival LSTM benchmarks while retaining explicit physical interpretability.
By Mohammad A. Farmani, Hoshin V. Gupta, Ali Behrangi, Muhammad Jawad, Sadaf Moghisi, Guo-Yue Niu
arXiv:2609.00396v1 Announce Type: new
Abstract: Histopathological whole slide images (WSIs) are central to cancer diagnosis, but their gigapixel scale, tissue heterogeneity, weak slide-level supervis...
By Chad Wong, Sicheng Chen, Tianyi Zhang, Enhui Chai, Yueming Jin, Zeyu Liu, Fei Xia
H2Table introduces a hierarchical hypergraph representation for complex tables, enabling a hypergraph encoder to capture semantic relationships between headers and cells. The framework uses learnable query vectors to extract structural embeddings for large language models. Experiments on the HiTab dataset show a 22.88% improvement over state‑of‑the‑art baselines on tables with four levels of nesting.
By Jia Ling, Yangfan Wang, Chen Tang, Haoming Tan, Yang Yang, Yi Guan, Jingchi Jiang
The paper introduces a second-pass method for uncovering hidden relationships in knowledge graphs extracted from text, without altering the original facts. By chunking documents and embedding each chunk once, the approach uses top‑k nearest‑neighbor queries and Shepard inverse‑distance weighting to score candidate node pairs, avoiding threshold issues inherent in cosine scoring. The technique is order‑independent, scalable, and has been implemented across multiple graph databases, demonstrating high edge fidelity with lower‑dimensional embeddings and a 25× speedup in top‑k computation.
By Bilge Kaan Karamete, Hunter Casten
SCoNE (Selective Context-aware Neuron Editing) is a training‑free method that enhances Retrieval‑Augmented Generation (RAG) by selectively strengthening context‑aware feed‑forward network neurons identified through high attribution and high cross‑input variability. It requires only a few mining samples, no fine‑tuning, and incurs no inference‑time overhead. Experiments on multiple knowledge‑intensive question‑answering benchmarks and two LLM backbones show that SCoNE consistently outperforms competitive baseline methods.
By Chaewon Kim, Seo Yeon Park