# Research Papers — 2025-10 · bundle page 209 of 217

> Full-text markdown of papers #10401–10450 of 10843 published in 2025-10, newest first. AI-readable mirrors of the full 10843-paper 2025-10 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2025-10/page/209
[← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/208.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-10/page/210.md)

### Attack logics, not outputs: Towards efficient robustification of deep neural networks by falsifying concept-based properties

Authors: Raik Dankworth, Gesina Schwalbe
Published: 2025-10-01
Categories: cs.CR, cs.LG
arXiv: https://arxiv.org/abs/2510.03320v1

Deep neural networks (NNs) for computer vision are vulnerable to adversarial attacks, i.e., miniscule malicious changes to inputs may induce unintuitive outputs. One key approach to verify and mitigate such robustness issues is to falsify expected output behavior. This allows, e.g., to locally proof security, or to (re)train NNs on obtained adversarial input examples. Due to the black-box nature of NNs, current attacks only falsify a class of the final output, such as flipping from $\texttt{stop_sign}$ to $\neg\texttt{stop_sign}$. In this short position paper we generalize this to search for generally illogical behavior, as considered in NN verification: falsify constraints (concept-based properties) involving further human-interpretable concepts, like $\texttt{red}\wedge\texttt{octogonal}\rightarrow\texttt{stop_sign}$. For this, an easy implementation of concept-based properties on already trained NNs is proposed using techniques from explainable artificial intelligence. Further, we sketch the theoretical proof that attacks on concept-based properties are expected to have a reduced search space compared to simple class falsification, whilst arguably be more aligned with intuitive robustness targets. As an outlook to this work in progress we hypothesize that this approach has potential to efficiently and simultaneously improve logical compliance and robustness.

---

### SVDefense: Effective Defense against Gradient Inversion Attacks via Singular Value Decomposition

Authors: Chenxiang Luo, David K. Y. Yau, Qun Song
Published: 2025-10-01
Categories: cs.CR, cs.LG
arXiv: https://arxiv.org/abs/2510.03319v1

Federated learning (FL) enables collaborative model training without sharing raw data but is vulnerable to gradient inversion attacks (GIAs), where adversaries reconstruct private data from shared gradients. Existing defenses either incur impractical computational overhead for embedded platforms or fail to achieve privacy protection and good model utility at the same time. Moreover, many defenses can be easily bypassed by adaptive adversaries who have obtained the defense details. To address these limitations, we propose SVDefense, a novel defense framework against GIAs that leverages the truncated Singular Value Decomposition (SVD) to obfuscate gradient updates. SVDefense introduces three key innovations, a Self-Adaptive Energy Threshold that adapts to client vulnerability, a Channel-Wise Weighted Approximation that selectively preserves essential gradient information for effective model training while enhancing privacy protection, and a Layer-Wise Weighted Aggregation for effective model aggregation under class imbalance. Our extensive evaluation shows that SVDefense outperforms existing defenses across multiple applications, including image classification, human activity recognition, and keyword spotting, by offering robust privacy protection with minimal impact on model accuracy. Furthermore, SVDefense is practical for deployment on various resource-constrained embedded platforms. We will make our code publicly available upon paper acceptance.

---

### Advances in Medical Image Segmentation: A Comprehensive Survey with a Focus on Lumbar Spine Applications

Authors: Ahmed Kabil, Ghada Khoriba, Mina Yousef, Essam A. Rashed
Published: 2025-10-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.03318v1

Medical Image Segmentation (MIS) stands as a cornerstone in medical image analysis, playing a pivotal role in precise diagnostics, treatment planning, and monitoring of various medical conditions. This paper presents a comprehensive and systematic survey of MIS methodologies, bridging the gap between traditional image processing techniques and modern deep learning approaches. The survey encompasses thresholding, edge detection, region-based segmentation, clustering algorithms, and model-based techniques while also delving into state-of-the-art deep learning architectures such as Convolutional Neural Networks (CNNs), Fully Convolutional Networks (FCNs), and the widely adopted U-Net and its variants. Moreover, integrating attention mechanisms, semi-supervised learning, generative adversarial networks (GANs), and Transformer-based models is thoroughly explored. In addition to covering established methods, this survey highlights emerging trends, including hybrid architectures, cross-modality learning, federated and distributed learning frameworks, and active learning strategies, which aim to address challenges such as limited labeled datasets, computational complexity, and model generalizability across diverse imaging modalities. Furthermore, a specialized case study on lumbar spine segmentation is presented, offering insights into the challenges and advancements in this relatively underexplored anatomical region. Despite significant progress in the field, critical challenges persist, including dataset bias, domain adaptation, interpretability of deep learning models, and integration into real-world clinical workflows.

---

### Photorealistic Inpainting for Perturbation-based Explanations in Ecological Monitoring

Authors: Günel Aghakishiyeva, Jiayi Zhou, Saagar Arya, Julian Dale, James David Poling, Holly R. Houliston, Jamie N. Womble, Gregory D. Larsen, David W. Johnston, Brinnae Bent
Published: 2025-10-01
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.03317v2

Ecological monitoring is increasingly automated by vision models, yet opaque predictions limit trust and field adoption. We present an inpainting-guided, perturbation-based explanation technique that produces photorealistic, mask-localized edits that preserve scene context. Unlike masking or blurring, these edits stay in-distribution and reveal which fine-grained morphological cues drive predictions in tasks such as species recognition and trait attribution. We demonstrate the approach on a YOLOv9 detector fine-tuned for harbor seal detection in Glacier Bay drone imagery, using Segment-Anything-Model-refined masks to support two interventions: (i) object removal/replacement (e.g., replacing seals with plausible ice/water or boats) and (ii) background replacement with original animals composited onto new scenes. Explanations are assessed by re-scoring perturbed images (flip rate, confidence drop) and by expert review for ecological plausibility and interpretability. The resulting explanations localize diagnostic structures, avoid deletion artifacts common to traditional perturbations, and yield domain-relevant insights that support expert validation and more trustworthy deployment of AI in ecology.

---

### The View From Space: Navigating Instrumentation Differences with EOFMs

Authors: Ryan P. Demilt, Nicholas LaHaye, Karis Tenneson
Published: 2025-10-01
Categories: cs.CV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.03316v1

Earth Observation Foundation Models (EOFMs) have exploded in prevalence as tools for processing the massive volumes of remotely sensed and other earth observation data, and for delivering impact on the many essential earth monitoring tasks. An emerging trend posits using the outputs of pre-trained models as 'embeddings' which summarize high dimensional data to be used for generic tasks such as similarity search and content-specific queries. However, most EOFM models are trained only on single modalities of data and then applied or benchmarked by matching bands across different modalities. It is not clear from existing work what impact diverse sensor architectures have on the internal representations of the present suite of EOFMs. We show in this work that the representation space of EOFMs is highly sensitive to sensor architecture and that understanding this difference gives a vital perspective on the pitfalls of current EOFM design and signals for how to move forward as model developers, users, and a community guided by robust remote-sensing science.

---

### Decomposing Attention To Find Context-Sensitive Neurons

Authors: Alex Gibson
Published: 2025-10-01
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.03315v1

We study transformer language models, analyzing attention heads whose attention patterns are spread out, and whose attention scores depend weakly on content. We argue that the softmax denominators of these heads are stable when the underlying token distribution is fixed. By sampling softmax denominators from a "calibration text", we can combine together the outputs of multiple such stable heads in the first layer of GPT2-Small, approximating their combined output by a linear summary of the surrounding text. This approximation enables a procedure where from the weights alone - and a single calibration text - we can uncover hundreds of first layer neurons that respond to high-level contextual properties of the surrounding text, including neurons that didn't activate on the calibration text.

---

### Glaucoma Detection and Structured OCT Report Generation via a Fine-tuned Multimodal Large Language Model

Authors: Jalil Jalili, Yashraj Gavhane, Evan Walker, Anna Heinke, Christopher Bowd, Akram Belghith, Massimo A. Fazio, Christopher A. Girkin, C. Gustavo De Moraes, Jeffrey M. Liebmann, Sally L. Baxter, Robert N. Weinreb et al.
Published: 2025-10-01
Categories: q-bio.QM, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2510.02403v1

Objective: To develop an explainable multimodal large language model (MM-LLM) that (1) screens optic nerve head (ONH) OCT circle scans for quality and (2) generates structured clinical reports that include glaucoma diagnosis and sector-wise retinal nerve fiber layer (RNFL) thinning assessments. Design: Retrospective cohort study of 1,310 subjects contributing 43,849 Spectralis ONH OCT circle scans (1,331 glaucomatous and 867 healthy eyes) from the DIGS and ADAGES cohorts. Methods: A MM-LLM (Llama 3.2 Vision-Instruct model) was fine-tuned to generate clinical descriptions of OCT imaging data. Training data included paired OCT images and automatically generated, structured clinical reports that described global and sectoral RNFL thinning. Poor-quality scans were labeled as unusable and paired with a fixed refusal statement. The model was evaluated on a held-out test set for three tasks: quality assessment, glaucoma detection, and RNFL thinning classification across seven anatomical sectors. Evaluation metrics included accuracy, sensitivity, specificity, precision, and F1-score. Model description quality was also evaluated using standard text evaluation metrics. Results: The model achieved 0.90 accuracy and 0.98 specificity for quality triage. For glaucoma detection, accuracy was 0.86 (sensitivity 0.91, specificity 0.73, F1-score 0.91). RNFL thinning prediction accuracy ranged from 0.83 to 0.94, with highest performance in global and temporal sectors. Text generation scores showed strong alignment with reference reports (BLEU: 0.82; ROUGE-1: 0.94; ROUGE-2: 0.87; ROUGE-L: 0.92; BERTScore-F1: 0.99). Conclusions: The fine-tuned MM-LLM generated accurate clinical descriptions based on OCT imaging. The model achieved high accuracy in identifying image quality issues and detecting glaucoma. The model also provided sectoral descriptions of RNFL thinning to help support clinical OCT evaluation.

---

### Linear RNNs for autoregressive generation of long music samples

Authors: Konrad Szewczyk, Daniel Gallo Fernández, James Townsend
Published: 2025-10-01
Categories: cs.SD, cs.AI, cs.LG, eess.AS
arXiv: https://arxiv.org/abs/2510.02401v1

Directly learning to generate audio waveforms in an autoregressive manner is a challenging task, due to the length of the raw sequences and the existence of important structure on many different timescales. Traditional approaches based on recurrent neural networks, as well as causal convolutions and self-attention, have only had limited success on this task. However, recent work has shown that deep state space models, also referred to as linear RNNs, can be highly efficient in this context. In this work, we push the boundaries of linear RNNs applied to raw audio modeling, investigating the effects of different architectural choices and using context-parallelism to enable training on sequences up to one minute (1M tokens) in length. We present a model, HarmonicRNN, which attains state of the art log-likelihoods and perceptual metrics on small-scale datasets.

---

### Retrieval and Augmentation of Domain Knowledge for Text-to-SQL Semantic Parsing

Authors: Manasi Patwardhan, Ayush Agarwal, Shabbirhussain Bhaisaheb, Aseem Arora, Lovekesh Vig, Sunita Sarawagi
Published: 2025-10-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.02394v1

The performance of Large Language Models (LLMs) for translating Natural Language (NL) queries into SQL varies significantly across databases (DBs). NL queries are often expressed using a domain specific vocabulary, and mapping these to the correct SQL requires an understanding of the embedded domain expressions, their relationship to the DB schema structure. Existing benchmarks rely on unrealistic, ad-hoc query specific textual hints for expressing domain knowledge. In this paper, we propose a systematic framework for associating structured domain statements at the database level. We present retrieval of relevant structured domain statements given a user query using sub-string level match. We evaluate on eleven realistic DB schemas covering diverse domains across five open-source and proprietary LLMs and demonstrate that (1) DB level structured domain statements are more practical and accurate than existing ad-hoc query specific textual domain statements, and (2) Our sub-string match based retrieval of relevant domain statements provides significantly higher accuracy than other retrieval approaches.

---

### KnowledgeSmith: Uncovering Knowledge Updating in LLMs with Model Editing and Unlearning

Authors: Yinyi Luo, Zhexian Zhou, Hao Chen, Kai Qiu, Marios Savvides, Sharon Li, Jindong Wang
Published: 2025-10-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.02392v3

Knowledge editing and machine unlearning are two popular approaches for large language models (LLMs) to stay up-to-date. However, the knowledge updating mechanism of LLMs remains largely unexplored due to insufficient, isolated, and small-scale evaluation. For instance, are LLMs similar to humans in modifying certain knowledge? What differs editing and unlearning as training data increases? This paper proposes KnowledgeSmith, a unified framework to systematically understand the updating mechanism of LLMs. We first cast editing and unlearning as instances of one constrained optimization problem. Then, we propose an automatic dataset generator that provides structured interventions across multiple graph levels and data scales, enabling controlled studies of how different modification strategies propagate through model knowledge. Extensive experiments demonstrate nuanced insights over knowledge propagation, plasticity scaling, consistency, and robustness. For instance, our results show that LLMs do not exhibit similar updating as humans for different levels of knowledge, and there exists consistency-capacity trade-off. We hope our findings can offer suggestions to the design of more reliable and scalable strategies. Code: https://github.com/AIFrontierLab/KnowledgeSmith.git

---

### Round-trip Reinforcement Learning: Self-Consistent Training for Better Chemical LLMs

Authors: Lecheng Kong, Xiyuan Wang, Yixin Chen, Muhan Zhang
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01527v2

Large Language Models (LLMs) are emerging as versatile foundation models for computational chemistry, handling bidirectional tasks like reaction prediction and retrosynthesis. However, these models often lack round-trip consistency. For instance, a state-of-the-art chemical LLM may successfully caption a molecule, yet be unable to accurately reconstruct the original structure from its own generated text. This inconsistency suggests that models are learning unidirectional memorization rather than flexible mastery. Indeed, recent work has demonstrated a strong correlation between a model's round-trip consistency and its performance on the primary tasks. This strong correlation reframes consistency into a direct target for model improvement. We therefore introduce Round-Trip Reinforcement Learning (RTRL), a novel framework that trains a model to improve its consistency by using the success of a round-trip transformation as a reward signal. We further propose an iterative variant where forward and reverse mappings alternately train each other in a self-improvement loop, a process that is highly data-efficient and notably effective with the massive amount of unlabelled data common in chemistry. Experiments demonstrate that RTRL significantly \textbf{boosts performance and consistency} over strong baselines across supervised, self-supervised, and synthetic data regimes. This work shows that round-trip consistency is not just a desirable property but a trainable objective, offering a new path toward more robust and reliable foundation models.

---

### One More Question is Enough, Expert Question Decomposition (EQD) Model for Domain Quantitative Reasoning

Authors: Mengyu Wang, Sotirios Sabanis, Miguel de Carvalho, Shay B. Cohen, Tiejun Ma
Published: 2025-10-01
Categories: cs.CL, q-fin.CP
arXiv: https://arxiv.org/abs/2510.01526v1

Domain-specific quantitative reasoning remains a major challenge for large language models (LLMs), especially in fields requiring expert knowledge and complex question answering (QA). In this work, we propose Expert Question Decomposition (EQD), an approach designed to balance the use of domain knowledge with computational efficiency. EQD is built on a two-step fine-tuning framework and guided by a reward function that measures the effectiveness of generated sub-questions in improving QA outcomes. It requires only a few thousand training examples and a single A100 GPU for fine-tuning, with inference time comparable to zero-shot prompting. Beyond its efficiency, EQD outperforms state-of-the-art domain-tuned models and advanced prompting strategies. We evaluate EQD in the financial domain, characterized by specialized knowledge and complex quantitative reasoning, across four benchmark datasets. Our method consistently improves QA performance by 0.6% to 10.5% across different LLMs. Our analysis reveals an important insight: in domain-specific QA, a single supporting question often provides greater benefit than detailed guidance steps.

---

### On Integer Programming for the Binarized Neural Network Verification Problem

Authors: Woojin Kim, James R. Luedtke
Published: 2025-10-01
Categories: cs.LG, math.OC
arXiv: https://arxiv.org/abs/2510.01525v1

Binarized neural networks (BNNs) are feedforward neural networks with binary weights and activation functions. In the context of using a BNN for classification, the verification problem seeks to determine whether a small perturbation of a given input can lead it to be misclassified by the BNN, and the robustness of the BNN can be measured by solving the verification problem over multiple inputs. The BNN verification problem can be formulated as an integer programming (IP) problem. However, the natural IP formulation is often challenging to solve due to a large integrality gap induced by big-$M$ constraints. We present two techniques to improve the IP formulation. First, we introduce a new method for obtaining a linear objective for the multi-class setting. Second, we introduce a new technique for generating valid inequalities for the IP formulation that exploits the recursive structure of BNNs. We find that our techniques enable verifying BNNs against a higher range of input perturbation than existing IP approaches within a limited time.

---

### WALT: Web Agents that Learn Tools

Authors: Viraj Prabhu, Yutong Dai, Matthew Fernandez, Jing Gu, Krithika Ramakrishnan, Yanqi Luo, Silvio Savarese, Caiming Xiong, Junnan Li, Zeyuan Chen, Ran Xu
Published: 2025-10-01
Categories: cs.CV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.01524v1

Web agents promise to automate complex browser tasks, but current methods remain brittle -- relying on step-by-step UI interactions and heavy LLM reasoning that break under dynamic layouts and long horizons. Humans, by contrast, exploit website-provided functionality through high-level operations like search, filter, and sort. We introduce WALT (Web Agents that Learn Tools), a framework that reverse-engineers latent website functionality into reusable invocable tools. Rather than hypothesizing ad-hoc skills, WALT exposes robust implementations of automations already designed into websites -- spanning discovery (search, filter, sort), communication (post, comment, upvote), and content management (create, edit, delete). Tools abstract away low-level execution: instead of reasoning about how to click and type, agents simply call search(query) or create(listing). This shifts the computational burden from fragile step-by-step reasoning to reliable tool invocation. On VisualWebArena and WebArena, WALT achieves higher success with fewer steps and less LLM-dependent reasoning, establishing a robust and generalizable paradigm for browser automation.

---

### CarbonX: An Open-Source Tool for Computational Decarbonization Using Time Series Foundation Models

Authors: Diptyaroop Maji, Kang Yang, Prashant Shenoy, Ramesh K Sitaraman, Mani Srivastava
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01521v2

Computational decarbonization aims to reduce carbon emissions in computing and societal systems such as data centers, transportation, and built environments. This requires accurate, fine-grained carbon intensity forecasts, yet existing tools have several key limitations: (i) they require grid-specific electricity mix data, restricting use where such information is unavailable; (ii) they depend on separate grid-specific models that make it challenging to provide global coverage; and (iii) they provide forecasts without uncertainty estimates, limiting reliability for downstream carbon-aware applications. In this paper, we present CarbonX, an open-source tool that leverages Time Series Foundation Models (TSFMs) for a range of decarbonization tasks. CarbonX utilizes the versatility of TSFMs to provide strong performance across multiple tasks, such as carbon intensity forecasting and imputation, and across diverse grids. Using only historical carbon intensity data and a single general model, our tool achieves a zero-shot forecasting Mean Absolute Percentage Error (MAPE) of 15.82% across 214 grids worldwide. Across 13 benchmark grids, CarbonX performance is comparable with the current state-of-the-art, with an average MAPE of 9.59% and tail forecasting MAPE of 16.54%, while also providing prediction intervals with 95% coverage. CarbonX can provide forecasts for up to 21 days with minimal accuracy degradation. Further, when fully fine-tuned, CarbonX outperforms the statistical baselines by 1.2--3.9X on the imputation task. Overall, these results demonstrate that CarbonX can be used easily on any grid with limited data and still deliver strong performance, making it a practical tool for global-scale decarbonization.

---

### Predictive Modeling and Explainable AI for Veterinary Safety Profiles, Residue Assessment, and Health Outcomes Using Real-World Data and Physicochemical Properties

Authors: Hossein Sholehrasa, Xuan Xu, Doina Caragea, Jim E. Riviere, Majid Jaberi-Douraki
Published: 2025-10-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.01520v1

The safe use of pharmaceuticals in food-producing animals is vital to protect animal welfare and human food safety. Adverse events (AEs) may signal unexpected pharmacokinetic or toxicokinetic effects, increasing the risk of violative residues in the food chain. This study introduces a predictive framework for classifying outcomes (Death vs. Recovery) using ~1.28 million reports (1987-2025 Q1) from the U.S. FDA's OpenFDA Center for Veterinary Medicine. A preprocessing pipeline merged relational tables and standardized AEs through VeDDRA ontologies. Data were normalized, missing values imputed, and high-cardinality features reduced; physicochemical drug properties were integrated to capture chemical-residue links. We evaluated supervised models, including Random Forest, CatBoost, XGBoost, ExcelFormer, and large language models (Gemma 3-27B, Phi 3-12B). Class imbalance was addressed, such as undersampling and oversampling, with a focus on prioritizing recall for fatal outcomes. Ensemble methods(Voting, Stacking) and CatBoost performed best, achieving precision, recall, and F1-scores of 0.95. Incorporating Average Uncertainty Margin (AUM)-based pseudo-labeling of uncertain cases improved minority-class detection, particularly in ExcelFormer and XGBoost. Interpretability via SHAP identified biologically plausible predictors, including lung, heart, and bronchial disorders, animal demographics, and drug physicochemical properties. These features were strongly linked to fatal outcomes. Overall, the framework shows that combining rigorous data engineering, advanced machine learning, and explainable AI enables accurate, interpretable predictions of veterinary safety outcomes. The approach supports FARAD's mission by enabling early detection of high-risk drug-event profiles, strengthening residue risk assessment, and informing regulatory and clinical decision-making.

---

### From Videos to Indexed Knowledge Graphs -- Framework to Marry Methods for Multimodal Content Analysis and Understanding

Authors: Basem Rizk, Joel Walsh, Mark Core, Benjamin Nye
Published: 2025-10-01
Categories: cs.CV, cs.AI, cs.CL, cs.IR
arXiv: https://arxiv.org/abs/2510.01513v1

Analysis of multi-modal content can be tricky, computationally expensive, and require a significant amount of engineering efforts. Lots of work with pre-trained models on static data is out there, yet fusing these opensource models and methods with complex data such as videos is relatively challenging. In this paper, we present a framework that enables efficiently prototyping pipelines for multi-modal content analysis. We craft a candidate recipe for a pipeline, marrying a set of pre-trained models, to convert videos into a temporal semi-structured data format. We translate this structure further to a frame-level indexed knowledge graph representation that is query-able and supports continual learning, enabling the dynamic incorporation of new domain-specific knowledge through an interactive medium.

---

### Flock: A Knowledge Graph Foundation Model via Learning on Random Walks

Authors: Jinwoo Kim, Xingyue Huang, Krzysztof Olejniczak, Kyungbin Min, Michael Bronstein, Seunghoon Hong, İsmail İlkan Ceylan
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01510v3

We study the problem of zero-shot link prediction on knowledge graphs (KGs), which requires models to generalize to novel entities and novel relations. Knowledge graph foundation models (KGFMs) address this task by enforcing equivariance over both nodes and relations, which enables them to learn structural properties of nodes and relations that transfer to novel KGs with similar structure. However, the conventional notion of deterministic equivariance inherently limits the expressive power of KGFMs, as it prevents them from distinguishing relations that are structurally similar but semantically distinct. To overcome this limitation, we propose to leverage probabilistic node-relation equivariance, which preserves equivariance in distribution while using structured randomness to break symmetries at inference time. Building on this principle, we present Flock, a KGFM that iteratively samples random walks, encodes them into sequences, embeds them with a sequence model, and aggregates node and relation representations through learned pooling. Flock respects probabilistic node-relation equivariance and, crucially, is a universal approximator for isomorphism-invariant link-level functions over KGs. Empirically, Flock perfectly solves our new diagnostic dataset Petals on which current KGFMs fail, and achieves state-of-the-art performance on entity and relation prediction tasks across 54 KGs from diverse domains. Code is available at https://github.com/jw9730/flock.

---

### Realistic CDSS Drug Dosing with End-to-end Recurrent Q-learning for Dual Vasopressor Control

Authors: Will Y. Zou, Jean Feng, Alexandre Kalimouttou, Jennifer Yuntong Zhang, Christopher W. Seymour, Romain Pirracchio
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01508v2

Reinforcement learning (RL) applications in Clinical Decision Support Systems (CDSS) frequently encounter skepticism because models may recommend inoperable dosing decisions. We propose an end-to-end offline RL framework for dual vasopressor administration in Intensive Care Units (ICUs) that directly addresses this challenge through principled action space design. Our method integrates discrete, continuous, and directional dosing strategies with conservative Q-learning and incorporates a novel recurrent modeling using a replay buffer to capture temporal dependencies in ICU time-series data. Our comparative analysis of norepinephrine dosing strategies across different action space formulations reveals that the designed action spaces improve interpretability and facilitate clinical adoption while preserving efficacy. Empirical results on eICU and MIMIC demonstrate that action space design profoundly influences learned behavioral policies. Compared with baselines, the proposed methods achieve more than 3x expected reward improvements, while aligning with established clinical protocols.

---

### Behavioral Geometric Supervision Aligns Video Foundation Models with Human Social Perception

Authors: Kathy Garcia, Leyla Isik
Published: 2025-10-01
Categories: q-bio.NC, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.01502v2

Current video foundation models, including the strongest self-supervised models such as V-JEPA2, fail to capture how humans organize social information in dynamic scenes. For example, across a range of diverse vision models tested, none were able to predict human similarity judgments to social video clips as well as a sentence embedding model of the caption text (MPNet). We show this gap in vision model performance can be closed by a compact behavioral supervisory signal. We introduce behavioral geometric supervision (BGS): a hybrid objective that constrains local and global pairwise embedding geometry to match the relational similarity structure across videos. We apply this method using a new human similarity dataset, containing 49,484 odd-one-out judgments from 250 naturalistic social video clips, and low-rank adaptation across four ViT backbones (V-JEPA 2/2.1, TimeSformer, VideoMAE, and CLIP). We find that one of the best fine-tuned models, V-JEPA 2.1, nearly triples in performance compared to the pre-trained baseline and reaches close to the noise ceiling, exceeding the strongest sentence-embedding baseline. In addition, finetuned models (i) capture unique variance in human judgments that caption-based language embeddings do not, (ii) develop interpretable social-affective attributes (valence, arousal, and dominance) despite never being trained on any of these attributes, (iii) zero-shot transfer to a separate dataset of out-of-distribution abstract social interactions, and (iv) shift spatial attention from scene context to socially informative regions (faces, gaze, and interacting bodies). A matched language-distillation control fails to reproduce these gains, ruling out caption transfer as the mechanism. Our results show how a modest amount of human behavioral data can steer video models toward human-like social visual understanding.

---

### Lateral Tree-of-Thoughts Surpasses ToT by Incorporating Logically-Consistent, Low-Utility Candidates

Authors: Abhinav Madahar
Published: 2025-10-01
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.01500v1

Modern deployments increasingly allocate large test-time compute (thousands of tokens or many node expansions) to boost reliability. Under such budgets, standard Tree-of-Thoughts-style search exhibits two pathologies: breadth saturation (additional samples mostly produce near-duplicates, so width stops growing) and depth myopia (noisy short-horizon utilities prune branches whose payoff appears after a few more steps). We propose Lateral Tree-of-Thoughts (LToT), a drop-in controller that separates utility from logical consistency and treats low-utility but consistent candidates as assets rather than waste. The frontier is split into mainlines (high-utility candidates used for exploitation) and laterals (consistent, initially low-utility candidates that receive short, cheap probes before judgment). LToT explores laterals via Lateral Racing with Short-Circuit (LR--SC): a capped successive-halving race that spreads tiny probes across a very wide lateral set, uses width-aware thresholds with repeat-to-confirm, and immediately promotes a branch once its envelope clears the mainline bar; mainlines are kept intentionally narrow so surplus compute is invested where width is cheap. We prove a pseudolinear lateral cost $Θ(N_0 \log_η N_0)$ with logarithmically many rungs (initial lateral width $N_0$; culling factor $η>1$), in contrast to the exponential growth of uncapped mainlines. Empirical evaluations on benchmark tasks are in preparation and will be added in a future revision. In short, LToT turns large test-time budgets into principled diversity while preserving promotion discipline, mitigating saturation and myopia without inflating compute.

---

### Beyond Majority Voting: LLM Aggregation by Leveraging Higher-Order Information

Authors: Rui Ai, Yuqi Pan, David Simchi-Levi, Milind Tambe, Haifeng Xu
Published: 2025-10-01
Categories: cs.LG, cs.AI, cs.GT
arXiv: https://arxiv.org/abs/2510.01499v2

With the rapid progress of multi-agent large language model (LLM) reasoning, how to effectively aggregate answers from multiple LLMs has emerged as a fundamental challenge. Standard majority voting treats all answers equally, failing to consider latent heterogeneity and correlation across models. In this work, we design two new aggregation algorithms called Optimal Weight (OW) and Inverse Surprising Popularity (ISP), leveraging both first-order and second-order information. Our theoretical analysis shows these methods provably mitigate inherent limitations of majority voting under mild assumptions, leading to more reliable collective decisions. We empirically validate our algorithms on synthetic datasets, popular LLM fine-tuning benchmarks such as UltraFeedback and MMLU, and a real-world healthcare setting ARMMAN. Our algorithms consistently outperform standard baselines, establishing a robust, training-free framework for effective multi-agent LLM aggregation.

---

### AortaDiff: A Unified Multitask Diffusion Framework For Contrast-Free AAA Imaging

Authors: Yuxuan Ou, Ning Bi, Jiazhen Pan, Jiancheng Yang, Boliang Yu, Usama Zidan, Regent Lee, Vicente Grau
Published: 2025-10-01
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.01498v2

While contrast-enhanced CT (CECT) is standard for assessing abdominal aortic aneurysms (AAA), the required iodinated contrast agents pose significant risks, including nephrotoxicity, patient allergies, and environmental harm. To reduce contrast agent use, recent deep learning methods have focused on generating synthetic CECT from non-contrast CT (NCCT) scans. However, most adopt a multi-stage pipeline that first generates images and then performs segmentation, which leads to error accumulation and fails to leverage shared semantic and anatomical structures. To address this, we propose a unified deep learning framework that generates synthetic CECT images from NCCT scans while simultaneously segmenting the aortic lumen and thrombus. Our approach integrates conditional diffusion models (CDM) with multi-task learning, enabling end-to-end joint optimization of image synthesis and anatomical segmentation. Unlike previous multitask diffusion models, our approach requires no initial predictions (e.g., a coarse segmentation mask), shares both encoder and decoder parameters across tasks, and employs a semi-supervised training strategy to learn from scans with missing segmentation labels, a common constraint in real-world clinical data. We evaluated our method on a cohort of 264 patients, where it consistently outperformed state-of-the-art single-task and multi-stage models. For image synthesis, our model achieved a PSNR of 25.61 dB, compared to 23.80 dB from a single-task CDM. For anatomical segmentation, it improved the lumen Dice score to 0.89 from 0.87 and the challenging thrombus Dice score to 0.53 from 0.48 (nnU-Net). These segmentation enhancements led to more accurate clinical measurements, reducing the lumen diameter MAE to 4.19 mm from 5.78 mm and the thrombus area error to 33.85% from 41.45% when compared to nnU-Net. Code is available at https://github.com/yuxuanou623/AortaDiff.git.

---

### Understanding Adversarial Transfer: Why Representation-Space Attacks Fail Where Data-Space Attacks Succeed

Authors: Isha Gupta, Rylan Schaeffer, Joshua Kazdan, Ken Ziyu Liu, Sanmi Koyejo
Published: 2025-10-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.01494v2

The field of adversarial robustness has long established that adversarial examples can successfully transfer between image classifiers and that text jailbreaks can successfully transfer between language models (LMs). However, a pair of recent studies reported being unable to successfully transfer image jailbreaks between vision-language models (VLMs). To explain this striking difference, we propose a fundamental distinction regarding the transferability of attacks against machine learning models: attacks in the input data-space can transfer, whereas attacks in model representation space do not, at least not without geometric alignment of representations. We then provide theoretical and empirical evidence of this hypothesis in four different settings. First, we mathematically prove this distinction in a simple setting where two networks compute the same input-output map but via different representations. Second, we construct representation-space attacks against image classifiers that are as successful as well-known data-space attacks, but fail to transfer. Third, we construct representation-space attacks against LMs that successfully jailbreak the attacked models but again fail to transfer. Fourth, we construct data-space attacks against VLMs that successfully transfer to new VLMs, and we show that representation space attacks can transfer when VLMs' latent geometries are sufficiently aligned in post-projector space. Our work reveals that adversarial transfer is not an inherent property of all attacks but contingent on their operational domain - the shared data-space versus models' unique representation spaces - a critical insight for building more robust models.

---

### Spatiotemporal Knowledge Graphs as Persistent Scene Memory for Embodied Question Answering

Authors: Mohamad Al Mdfaa, Svetlana Lukina, Timur Akhtyamov, Arthur Nigmatzyanov, Dmitrii Nalberskii, Sergey Zagoruyko, Gonzalo Ferrer
Published: 2025-10-01
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2510.01483v3

Vision-language models (VLMs) demonstrate strong image-level scene understanding, but reasoning over long egocentric video remains costly: because VLMs maintain no persistent memory or explicit spatial representation, all sampled frames must be re-processed for every new query. We present VL-KnG, a training-free framework that constructs spatiotemporal knowledge graphs from monocular egocentric video, bridging fine-grained scene graphs and global topological graphs without 3D reconstruction. VL-KnG processes video in chunks, maintains persistent object identities via large language model (LLM)-based Spatiotemporal Object Association (STOA), and answers questions through Graph-Enhanced Retrieval (GER), which combines subgraph retrieval with visual grounding. Once constructed, the knowledge graph removes the need to revisit raw video at query time, decoupling query latency from video length. Across three embodied question answering benchmarks -- OpenEQA, NaVQA, and our newly introduced WalkieKnowledge -- VL-KnG achieves accuracy competitive with frontier VLMs while answering queries at substantially lower latency and with explainable, graph-grounded reasoning; it further surpasses prior persistent-representation baselines and open-weight VLMs in several settings. Deployment on a real robot demonstrates practical applicability, with query latency remaining stable as observation history grows. As a persistent, queryable scene memory, VL-KnG provides a concrete substrate for memory maintenance and knowledge updating in embodied AI agents.

---

### Pharmacophore-Guided Generative Design of Novel Drug-Like Molecules

Authors: Ekaterina Podplutova, Anastasia Vepreva, Olga A. Konovalova, Vladimir Vinogradov, Dmitrii O. Shkil, Andrei Dmitrenko
Published: 2025-10-01
Categories: q-bio.QM, cs.AI
arXiv: https://arxiv.org/abs/2510.01480v1

The integration of artificial intelligence (AI) in early-stage drug discovery offers unprecedented opportunities for exploring chemical space and accelerating hit-to-lead optimization. However, docking optimization in generative approaches is computationally expensive and may lead to inaccurate results. Here, we present a novel generative framework that balances pharmacophore similarity to reference compounds with structural diversity from active molecules. The framework allows users to provide custom reference sets, including FDA-approved drugs or clinical candidates, and guides the \textit{de novo} generation of potential therapeutics. We demonstrate its applicability through a case study targeting estrogen receptor modulators and antagonists for breast cancer. The generated compounds maintain high pharmacophoric fidelity to known active molecules while introducing substantial structural novelty, suggesting strong potential for functional innovation and patentability. Comprehensive evaluation of the generated molecules against common drug-like properties confirms the robustness and pharmaceutical relevance of the approach.

---

### Density-Ratio Weighted Behavioral Cloning: Learning Control Policies from Corrupted Datasets

Authors: Shriram Karpoora Sundara Pandian, Ali Baheri
Published: 2025-10-01
Categories: cs.LG, eess.SY
arXiv: https://arxiv.org/abs/2510.01479v2

Offline reinforcement learning (RL) enables policy optimization from fixed datasets, making it suitable for safety-critical applications where online exploration is infeasible. However, these datasets are often contaminated by adversarial poisoning, system errors, or low-quality samples, leading to degraded policy performance in standard behavioral cloning (BC) and offline RL methods. This paper introduces Density-Ratio Weighted Behavioral Cloning (Weighted BC), a robust imitation learning approach that uses a small, verified clean reference set to estimate trajectory-level density ratios via a binary discriminator. These ratios are clipped and used as weights in the BC objective to prioritize clean expert behavior while down-weighting or discarding corrupted data, without requiring knowledge of the contamination mechanism. We establish theoretical guarantees showing convergence to the clean expert policy with finite-sample bounds that are independent of the contamination rate. A comprehensive evaluation framework is established, which incorporates various poisoning protocols (reward, state, transition, and action) on continuous control benchmarks. Experiments demonstrate that Weighted BC maintains near-optimal performance even at high contamination ratios outperforming baselines such as traditional BC, batch-constrained Q-learning (BCQ) and behavior regularized actor-critic (BRAC).

---

### Purrception: Variational Flow Matching for Vector-Quantized Image Generation

Authors: Răzvan-Andrei Matişan, Vincent Tao Hu, Grigory Bartosh, Björn Ommer, Cees G. M. Snoek, Max Welling, Jan-Willem van de Meent, Mohammad Mahdi Derakhshani, Floor Eijkelboom
Published: 2025-10-01
Categories: cs.CV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.01478v4

We introduce Purrception, a variational flow matching approach for vector-quantized image generation that provides explicit categorical supervision while maintaining continuous transport dynamics. Our method adapts Variational Flow Matching to vector-quantized latents by learning categorical posteriors over codebook indices while computing velocity fields in the continuous embedding space. This combines the geometric awareness of continuous methods with the discrete supervision of categorical approaches, enabling uncertainty quantification over plausible codes and temperature-controlled generation. We evaluate Purrception on ImageNet-1k 256x256 generation. Training converges faster than both continuous flow matching and discrete flow matching baselines while achieving competitive FID scores with state-of-the-art models. This demonstrates that Variational Flow Matching can effectively bridge continuous transport and discrete supervision for improved training efficiency in image generation.

---

### Comparative Field Deployment of Reinforcement Learning and Model Predictive Control for Residential HVAC

Authors: Ozan Baris Mulayim, Elias N. Pergantis, Levi D. Reyes Premer, Bingqing Chen, Guannan Qu, Kevin J. Kircher, Mario Bergés
Published: 2025-10-01
Categories: eess.SY, cs.LG
arXiv: https://arxiv.org/abs/2510.01475v2

Model Predictive Control (MPC) has demonstrated significant performance improvements over today's control methods for residential Heating, Ventilation, and Air Conditioning (HVAC), but deploying MPC often requires substantial engineering effort. Reinforcement Learning (RL) may offer comparable performance with easier deployment, but its practical application for residential HVAC remains largely undemonstrated, leaving open questions related to occupant comfort and data requirements. To investigate these issues, we deployed one MPC variant and one model-based RL variant for one month each in an occupied house in a cold climate. The controllers adjusted an air-to-air heat pump's thermostat temperature setpoint based on measurements of the indoor temperature and the electric power used for heating. Relative to constant-setpoint operation, MPC saved 18.1\% (95\% confidence interval: 4.4 to 30.9\%) of weather-normalized heat pump energy and RL saved 20.9\% (2.6 to 38.3\%). MPC maintained acceptable occupant comfort. RL kept the house cooler, particularly during an initial adaptation phase, leading to three reports of occupant discomfort. The two algorithms had similar data requirements. We estimate that for a fresh deployment in another house, RL would take about one-third less engineering effort than MPC. While RL reduces deployment effort, it faces difficulties related to safe controller initialization and to mismatches between the modeled and true state and action spaces.

---

### AIReg-Bench: Benchmarking Language Models That Assess AI Regulation Compliance

Authors: Bill Marino, Rosco Hunter, Christoph Schnabl, Zubair Jamali, Marinos Emmanouil Kalpakos, Mudra Kashyap, Isaiah Hinton, Alexa Hanson, Maahum Nazir, Felix Steffek, Hongkai Wen, Nicholas D. Lane
Published: 2025-10-01
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.01474v3

As governments move to regulate AI, there is growing interest in using Large Language Models (LLMs) to assess whether or not an AI system complies with a given AI Regulation (AIR). However, there is presently no way to benchmark the performance of LLMs at this task. To fill this void, we introduce AIReg-Bench: the first open benchmark dataset designed to test how well LLMs can assess compliance with the EU AI Act (AIA). We created this dataset through a two-step process: (1) by prompting an LLM with carefully structured instructions, we generated 120 technical documentation excerpts (samples), each depicting a fictional, albeit plausible, AI system -- of the kind an AI provider might produce to demonstrate their compliance with AIR; (2) legal experts then reviewed and annotated each sample to indicate whether, and in what way, the AI system described therein violates specific Articles of the AIA. The resulting dataset, together with our evaluation of whether frontier LLMs can reproduce the experts' compliance labels, provides a starting point to understand the opportunities and limitations of LLM-based AIR compliance assessment tools and establishes a benchmark against which subsequent LLMs can be compared. The dataset and evaluation code are available at https://github.com/camlsys/aireg-bench.

---

### From keywords to semantics: Perceptions of large language models in data discovery

Authors: Maura E Halstead, Mark A. Green, Caroline Jay, Richard Kingston, David Topping, Alexander Singleton
Published: 2025-10-01
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2510.01473v1

Current approaches to data discovery match keywords between metadata and queries. This matching requires researchers to know the exact wording that other researchers previously used, creating a challenging process that could lead to missing relevant data. Large Language Models (LLMs) could enhance data discovery by removing this requirement and allowing researchers to ask questions with natural language. However, we do not currently know if researchers would accept LLMs for data discovery. Using a human-centered artificial intelligence (HCAI) focus, we ran focus groups (N = 27) to understand researchers' perspectives towards LLMs for data discovery. Our conceptual model shows that the potential benefits are not enough for researchers to use LLMs instead of current technology. Barriers prevent researchers from fully accepting LLMs, but features around transparency could overcome them. Using our model will allow developers to incorporate features that result in an increased acceptance of LLMs for data discovery.

---

### LLM-NAS: LLM-driven Hardware-Aware Neural Architecture Search

Authors: Hengyi Zhu, Grace Li Zhang, Shaoyi Huang
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01472v4

Hardware-Aware Neural Architecture Search (HW-NAS) requires joint optimization of accuracy and latency under device constraints. Traditional supernet-based methods require multiple GPU days per dataset. Large Language Model (LLM)-driven approaches avoid training a large supernet and can provide quick feedback, but we observe an exploration bias: the LLM repeatedly proposes neural network designs within limited search space and fails to discover architectures across different latency ranges in the entire search space. To address this issue, we propose LLM-NAS: an LLM-driven Neural Architecture Search that can generate neural networks with high accuracy and low latency with reduced search cost. Our proposed LLM-NAS has three key components: 1) a complexity-driven partitioning engine that divides the search space by complexity to enforce diversity and mitigate exploration bias; 2) an LLM-powered architecture prompt co-evolution operator, in which the LLM first updates a knowledge base of design heuristics based on results from the previous round, then performs a guided evolution algorithm on architectures with prompts that incorporate this knowledge base. Prompts and designs improve together across rounds which avoids random guesswork and improve efficiency; 3) a zero-cost predictor to avoid training a large number of candidates from scratch. Experimental results show that on HW-NAS-Bench, LLM-NAS can achieve overall higher HV, lower IGD, and up to 54% lower latency than baselines at similar accuracy. Meanwhile, the search cost drops from days to minutes compared with traditional supernet baselines.

---

### Fine-tuning LLMs with variational Bayesian last layer for high-dimensional Bayesian optimization

Authors: Haotian Xiang, Jinwen Xu, Qin Lu
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01471v2

A plethora of applications entail solving black-box optimization problems with high evaluation costs, including drug discovery, material design, as well as hyperparameter tuning. Toward finding the global optimum of such black-box optimization problems with sample efficiency, Bayesian optimization (BO) is a theoretically elegant framework that relies on a probabilistic surrogate model so as to iteratively select the query point with well-balanced exploration-exploitation tradeoffs. The Gaussian process (GP), as the de-facto choice for surrogate modeling, has achieved compelling performances for vanilla BO with low-dimensional continuous variables. However, GPs fall short in coping with high-dimensional counterparts with {\it irregular} variables (e.g., categorical, ordinal, etc.). To alleviate this, neural network-based surrogates have been explored. Inspired by the powerful capabilities of LLMs, we adopt the LLM as the surrogate to model the mapping from the high-dimensional input variables to the objective function. To adapt to the current problem, we leverage the low-rank adaptation (LoRA) to fine-tune the LLM parameters together with the posterior of a linear regression head via the variational Bayesian last layer (VBLL) framework. The resulting LoRA-VBLL is not only computationally light compared to existing alternatives, but also admits recursive updates. To automate the critical selection of the LoRA rank as well as other hyperparameters, a weighted ensemble (ENS) of LoRA-VBLL surrogates has been devised, which further accommodates continual update of the per-model weight and individual LoRA-VBLL parameters via recursive Bayes. Extensive experimental results demonstrate the compelling performance of the proposed (ENS-)LoRA-VBLL approaches on various high-dimensional benchmarks and the real-world molecular optimization tasks.

---

### Extracting O*NET Features from the NLx Corpus to Build Public Use Aggregate Labor Market Data

Authors: Stephen Meisenbacher, Svetlozar Nestorov, Peter Norlander
Published: 2025-10-01
Categories: cs.CY, cs.CL
arXiv: https://arxiv.org/abs/2510.01470v1

Data from online job postings are difficult to access and are not built in a standard or transparent manner. Data included in the standard taxonomy and occupational information database (O*NET) are updated infrequently and based on small survey samples. We adopt O*NET as a framework for building natural language processing tools that extract structured information from job postings. We publish the Job Ad Analysis Toolkit (JAAT), a collection of open-source tools built for this purpose, and demonstrate its reliability and accuracy in out-of-sample and LLM-as-a-Judge testing. We extract more than 10 billion data points from more than 155 million online job ads provided by the National Labor Exchange (NLx) Research Hub, including O*NET tasks, occupation codes, tools, and technologies, as well as wages, skills, industry, and more features. We describe the construction of a dataset of occupation, state, and industry level features aggregated by monthly active jobs from 2015 - 2025. We illustrate the potential for research and future uses in education and workforce development.

---

### A-VERT: Agnostic Verification with Embedding Ranking Targets

Authors: Nicolás Aguirre, Ramiro Caso, Ramiro Rodríguez Colmeiro, Mauro Santelli, Joaquín Toranzo Calderón
Published: 2025-10-01
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2510.01469v1

The automatic evaluation of Language Model (LM) responses is a critical piece in the development of benchmarks and metrics, both for model training and quality assessment of production model endpoints. The current approaches to response classification relies on methods that are too expensive (i.e. LLM-as-a-Judge) or that are far from real-world conditions (string-matching, logprob). In this paper, a structure-free evaluation method is presented. The method makes use of semantic embedding distances to match target candidates with arbitrary LM-generated text, resulting in a robust classification of the response at a relatively low compute cost (embedding models of less than $10B$ parameters). The results show a regression score of ~0.97 and an accuracy of ~96% against human annotators, tested over 3 data sets and 3 different LM architectures.

---

### RealClass: A Framework for Classroom Speech Simulation with Public Datasets and Game Engines

Authors: Ahmed Adel Attia, Jing Liu, Carol Espy Wilson
Published: 2025-10-01
Categories: cs.SD, cs.AI, eess.AS
arXiv: https://arxiv.org/abs/2510.01462v1

The scarcity of large-scale classroom speech data has hindered the development of AI-driven speech models for education. Classroom datasets remain limited and not publicly available, and the absence of dedicated classroom noise or Room Impulse Response (RIR) corpora prevents the use of standard data augmentation techniques. In this paper, we introduce a scalable methodology for synthesizing classroom noise and RIRs using game engines, a versatile framework that can extend to other domains beyond the classroom. Building on this methodology, we present RealClass, a dataset that combines a synthesized classroom noise corpus with a classroom speech dataset compiled from publicly available corpora. The speech data pairs a children's speech corpus with instructional speech extracted from YouTube videos to approximate real classroom interactions in clean conditions. Experiments on clean and noisy speech show that RealClass closely approximates real classroom speech, making it a valuable asset in the absence of abundant real classroom speech.

---

### The Three Regimes of Offline-to-Online Reinforcement Learning

Authors: Lu Li, Tianwei Ni, Yihao Sun, Pierre-Luc Bacon
Published: 2025-10-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.01460v4

Offline-to-online reinforcement learning (RL) has emerged as a practical paradigm that leverages offline datasets for pretraining and online interactions for fine-tuning. However, its empirical behavior is highly inconsistent: design choices of online fine-tuning that work well in one setting can fail completely in another. Guided by the stability--plasticity principle, we propose a framework that can explain this inconsistency: We argue that efficient fine-tuning must preserve the utility of the stronger offline prior, whether that is the pretrained policy or the offline dataset, while maintaining sufficient plasticity. This perspective identifies three regimes of online fine-tuning, each requiring distinct stability properties. We validate this framework through a large-scale empirical study, finding that the results strongly align with its predictions in 45 out of 63 cases, with only 3 opposite mismatches. This work provides a framework for guiding design choices in offline-to-online RL based on the relative performance of the offline dataset and the pretrained policy.

---

### LSPO: Length-aware Dynamic Sampling for Policy Optimization in LLM Reasoning

Authors: Weizhe Chen, Sven Koenig, Bistra Dilkina
Published: 2025-10-01
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2510.01459v1

Since the release of Deepseek-R1, reinforcement learning with verifiable rewards (RLVR) has become a central approach for training large language models (LLMs) on reasoning tasks. Recent work has largely focused on modifying loss functions to make RLVR more efficient and effective. In this paper, motivated by studies of overthinking in LLMs, we propose Length-aware Sampling for Policy Optimization (LSPO), a novel meta-RLVR algorithm that dynamically selects training data at each step based on the average response length. We evaluate LSPO across multiple base models and datasets, demonstrating that it consistently improves learning effectiveness. In addition, we conduct a detailed ablation study to examine alternative ways of incorporating length signals into dynamic sampling, offering further insights and highlighting promising directions for future research.

---

### How Well Can Preference Optimization Generalize Under Noisy Feedback?

Authors: Shawn Im, Sharon Li
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01458v3

As large language models (LLMs) advance their capabilities, aligning these models with human preferences has become crucial. Preference optimization, which trains models to distinguish between preferred and non-preferred responses based on human feedback, has become a crucial component for aligning LLMs. However, most existing works assume noise-free feedback, which is unrealistic due to the inherent errors and inconsistencies in human judgments. This paper addresses the impact of noisy feedback on preference optimization, providing generalization guarantees under these conditions. In particular, we consider noise models that correspond to common real-world sources of noise, such as mislabeling and uncertainty. Unlike traditional analyses that assume convergence, our work focuses on finite-step preference optimization, offering new insights that are more aligned with practical LLM training. We describe how generalization decays with different types of noise across levels of noise rates based on the preference data distribution and number of samples. Our analysis for noisy preference learning applies to a broad family of preference optimization losses such as DPO, IPO, SLiC, etc. Empirical validation on contemporary LLMs confirms the practical relevance of our findings, offering valuable insights for developing AI systems that align with human preferences.

---

### A Forensic Analysis of Synthetic Data in RL: Diagnosing and Solving Algorithmic Failures in Model-Based Policy Optimization

Authors: Brett Barkley, David Fridovich-Keil
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01457v4

Synthetic data is central to data-efficient Dyna-style model-based reinforcement learning, but it can also degrade performance. We study this failure in Model-Based Policy Optimization (MBPO), which performs actor-critic updates using model-generated synthetic state transitions. Although MBPO reports strong sample-efficiency gains on OpenAI Gym, recent work shows that it often underperforms Soft Actor-Critic (SAC), its non-Dyna base, in the DeepMind Control Suite (DMC), despite both suites involving MuJoCo-based proprioceptive continuous control. We identify two coupled causes of this collapse: scale mismatch between dynamics and reward targets, which suppresses reward learning and induces critic underestimation, and residual next-state prediction, which inflates model variance and produces unreliable synthetic transitions. We introduce Fixing That Free Lunch (FTFL), a minimal repair that combines independent target normalization with direct next-state prediction. FTFL outperforms SAC in five of seven previously failing DMC tasks while preserving MBPO's strong Gym performance. We further show that MBPO-lineage algorithms, including uncertainty-aware variants that filter, penalize, or reject synthetic transitions based on model uncertainty, still inherit these failures unless FTFL is applied to their shared learned-model backbone. More broadly, our results show how benchmark-limited evaluation can encode environment-specific assumptions into algorithm design, motivating taxonomies that map MDP structure to algorithmic failure modes.

---

### SCOPED: Score-Curvature Out-of-distribution Proximity Evaluator for Diffusion

Authors: Brett Barkley, Preston Culbertson, David Fridovich-Keil
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01456v2

Out-of-distribution (OOD) detection is essential for reliable deployment of machine learning systems in vision, robotics, reinforcement learning, and beyond. We introduce Score-Curvature Out-of-distribution Proximity Evaluator for Diffusion (SCOPED), a fast and general-purpose OOD detection method for diffusion models that reduces the number of forward passes on the trained model by an order of magnitude compared to prior methods, outperforming most diffusion-based baselines and closely approaching the accuracy of the strongest ones. SCOPED is computed from a single diffusion model trained once on a diverse dataset, and combines the Jacobian trace and squared norm of the model's score function into a single test statistic. Rather than thresholding on a fixed value, we estimate the in-distribution density of SCOPED scores using kernel density estimation, enabling a flexible, unsupervised test that, in the simplest case, only requires a single forward pass and one Jacobian-vector product (JVP), made efficient by Hutchinson's trace estimator. On four vision benchmarks, SCOPED achieves competitive or state-of-the-art precision-recall scores despite its low computational cost. The same method generalizes to robotic control tasks with shared state and action spaces, identifying distribution shifts across reward functions and training regimes. These results position SCOPED as a practical foundation for fast and reliable OOD detection in real-world domains, including perceptual artifacts in vision, outlier detection in autoregressive models, exploration in reinforcement learning, and dataset curation for unsupervised training.

---

### Data Selection for Fine-tuning Vision Language Models via Cross Modal Alignment Trajectories

Authors: Nilay Naharas, Dang Nguyen, Nesihan Bulut, Mohammadhossein Bateni, Vahab Mirrokni, Baharan Mirzasoleiman
Published: 2025-10-01
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.01454v1

Data-efficient learning aims to eliminate redundancy in large training datasets by training models on smaller subsets of the most informative examples. While data selection has been extensively explored for vision models and large language models (LLMs), it remains underexplored for Large Vision-Language Models (LVLMs). Notably, none of existing methods can outperform random selection at different subset sizes. In this work, we propose the first principled method for data-efficient instruction tuning of LVLMs. We prove that examples with similar cross-modal attention matrices during instruction tuning have similar gradients. Thus, they influence model parameters in a similar manner and convey the same information to the model during training. Building on this insight, we propose XMAS, which clusters examples based on the trajectories of the top singular values of their attention matrices obtained from fine-tuning a small proxy LVLM. By sampling a balanced subset from these clusters, XMAS effectively removes redundancy in large-scale LVLM training data. Extensive experiments show that XMAS can discard 50% of the LLaVA-665k dataset and 85% of the Vision-Flan dataset while fully preserving performance of LLaVA-1.5-7B on 10 downstream benchmarks and speeding up its training by 1.2x. This is 30% more data reduction compared to the best baseline for LLaVA-665k. The project's website can be found at https://bigml-cs-ucla.github.io/XMAS-project-page/.

---

### The Command Line GUIde: Graphical Interfaces from Man Pages via AI

Authors: Saketh Ram Kasibatla, Kiran Medleri Hiremath, Raven Rothkopf, Sorin Lerner, Haijun Xia, Brian Hempel
Published: 2025-10-01
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2510.01453v1

Although birthed in the era of teletypes, the command line shell survived the graphical interface revolution of the 1980's and lives on in modern desktop operating systems. The command line provides access to powerful functionality not otherwise exposed on the computer, but requires users to recall textual syntax and carefully scour documentation. In contrast, graphical interfaces let users organically discover and invoke possible actions through widgets and menus. To better expose the power of the command line, we demonstrate a mechanism for automatically creating graphical interfaces for command line tools by translating their documentation (in the form of man pages) into interface specifications via AI. Using these specifications, our user-facing system, called GUIde, presents the command options to the user graphically. We evaluate the generated interfaces on a corpus of commands to show to what degree GUIde offers thorough graphical interfaces for users' real-world command line tasks.

---

### Financial Stability Implications of Generative AI: Taming the Animal Spirits

Authors: Anne Lundgaard Hansen, Seung Jung Lee
Published: 2025-10-01
Categories: q-fin.GN, cs.LG
arXiv: https://arxiv.org/abs/2510.01451v1

This paper investigates the impact of the adoption of generative AI on financial stability. We conduct laboratory-style experiments using large language models to replicate classic studies on herd behavior in trading decisions. Our results show that AI agents make more rational decisions than humans, relying predominantly on private information over market trends. Increased reliance on AI-powered trading advice could therefore potentially lead to fewer asset price bubbles arising from animal spirits that trade by following the herd. However, exploring variations in the experimental settings reveals that AI agents can be induced to herd optimally when explicitly guided to make profit-maximizing decisions. While optimal herding improves market discipline, this behavior still carries potential implications for financial stability. In other experimental variations, we show that AI agents are not purely algorithmic, but have inherited some elements of human conditioning and bias.

---

### Local Linear Attention: An Optimal Interpolation of Linear and Softmax Attention For Test-Time Regression

Authors: Yifei Zuo, Yutong Yin, Zhichen Zeng, Ang Li, Banghua Zhu, Zhaoran Wang
Published: 2025-10-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.01450v1

Transformer architectures have achieved remarkable success in various domains. While efficient alternatives to Softmax Attention have been widely studied, the search for more expressive mechanisms grounded in theoretical insight-even at greater computational cost-has been relatively underexplored. In this work, we bridge this gap by proposing Local Linear Attention (LLA), a novel attention mechanism derived from nonparametric statistics through the lens of test-time regression. First, we show that LLA offers theoretical advantages over Linear and Softmax Attention for associative memory via a bias-variance trade-off analysis. Next, we address its computational challenges and propose two memory-efficient primitives to tackle the $Θ(n^2 d)$ and $Θ(n d^2)$ complexity. We then introduce FlashLLA, a hardware-efficient, blockwise algorithm that enables scalable and parallel computation on modern accelerators. In addition, we implement and profile a customized inference kernel that significantly reduces memory overheads. Finally, we empirically validate the advantages and limitations of LLA on test-time regression, in-context regression, associative recall and state tracking tasks. Experiment results demonstrate that LLA effectively adapts to non-stationarity, outperforming strong baselines in test-time training and in-context learning, and exhibiting promising evidence for its scalability and applicability in large-scale models. Code is available at https://github.com/Yifei-Zuo/Flash-LLA.

---

### GeoSURGE: Geo-localization using Semantic Fusion with Hierarchy of Geographic Embeddings

Authors: Angel Daruna, Nicholas Meegan, Han-Pang Chiu, Supun Samarasekera, Rakesh Kumar
Published: 2025-10-01
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.01448v2

Worldwide visual geo-localization aims to determine the geographic location of an image anywhere on Earth using only its visual content. Despite recent progress, learning expressive representations of geographic space remains challenging due to the inherently low-dimensional nature of geographic coordinates. We formulate global geo-localization as aligning the visual representation of a query image with a learned geographic representation. Our approach explicitly models the world as a hierarchy of learned geographic embeddings, enabling a distributed and multi-scale representation of geographic space. In addition, we introduce a semantic fusion module that efficiently integrates appearance features with semantic segmentation through latent cross-attention, producing a more robust visual representation for localization. Experiments on five widely used geo-localization benchmarks demonstrate that our method achieves new state-of-the-art results on 22 of 25 reported metrics. Ablation studies show that these improvements are primarily driven by the proposed geographic representation and semantic fusion mechanism.

---

### SoftAdaClip: A Smooth Clipping Strategy for Fair and Private Model Training

Authors: Dorsa Soleymani, Ali Dadsetan, Frank Rudzicz
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01447v2

Differential privacy (DP) provides strong protection for sensitive data, but often reduces model performance and fairness, especially for underrepresented groups. One major reason is gradient clipping in DP-SGD, which can disproportionately suppress learning signals for minority subpopulations. Although adaptive clipping can enhance utility, it still relies on uniform hard clipping, which may restrict fairness. To address this, we introduce SoftAdaClip, a differentially private training method that replaces hard clipping with a smooth, tanh-based transformation to preserve relative gradient magnitudes while bounding sensitivity. We evaluate SoftAdaClip on various datasets, including MIMIC-III (clinical text), GOSSIS-eICU (structured healthcare), and Adult Income (tabular data). Our results show that SoftAdaClip reduces subgroup disparities by up to 87% compared to DP-SGD and up to 48% compared to Adaptive-DPSGD, and these reductions in subgroup disparities are statistically significant. These findings underscore the importance of integrating smooth transformations with adaptive mechanisms to achieve fair and private model training.

---

### Dual-Uncertainty Guided Policy Learning for Multimodal Reasoning

Authors: Rui Liu, Dian Yu, Tong Zheng, Runpeng Dai, Zongxia Li, Wenhao Yu, Zhenwen Liang, Linfeng Song, Haitao Mi, Pratap Tokekar, Dong Yu
Published: 2025-10-01
Categories: cs.AI, cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2510.01444v3

Reinforcement learning with verifiable rewards (RLVR) has advanced reasoning capabilities in multimodal large language models. However, existing methods typically treat visual inputs as deterministic, overlooking the perceptual ambiguity inherent to the visual modality. Consequently, they fail to distinguish whether a model's uncertainty stems from complex reasoning or ambiguous perception, preventing the targeted allocation of exploration or learning signals. To address this gap, we introduce \textbf{DUPL}, a dual-uncertainty guided policy learning approach for multimodal RLVR that quantifies and leverages both perceptual uncertainty (via symmetric KL divergence) and output uncertainty (via policy entropy) to guide policy updates. By establishing an uncertainty-driven feedback loop and employing a dynamic branch prioritization mechanism, DUPL recalibrates the policy advantage to focus learning on states with high perceptual or decisional ambiguity, enabling effective targeted exploration beyond passive data augmentation. Evaluated on diverse multimodal reasoning benchmarks spanning mathematical and general domains, DUPL achieves solid gains. It improves Qwen2.5-VL accuracy by up to $\textbf{12.3%}$ (3B) and $\textbf{7.9%}$ (7B), and Qwen3-VL-Instruct by up to $\textbf{10.7%}$ (4B) and $\textbf{12.4%}$ (8B), consistently outperforming GRPO, while seamlessly generalizing to alternative algorithms (DAPO, $\textbf{+6.5%}$ avg) and architectures (LLaVA-OneVision-1.5, $\textbf{+4.7%}$ avg). These results demonstrate that DUPL is an effective and generalizable approach for multimodal RLVR.

---

### Edge Artificial Intelligence: A Systematic Review of Evolution, Taxonomic Frameworks, and Future Horizons

Authors: Mohamad Abou Ali, Fadi Dornaika
Published: 2025-10-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.01439v1

Edge Artificial Intelligence (Edge AI) embeds intelligence directly into devices at the network edge, enabling real-time processing with improved privacy and reduced latency by processing data close to its source. This review systematically examines the evolution, current landscape, and future directions of Edge AI through a multi-dimensional taxonomy including deployment location, processing capabilities such as TinyML and federated learning, application domains, and hardware types. Following PRISMA guidelines, the analysis traces the field from early content delivery networks and fog computing to modern on-device intelligence. Core enabling technologies such as specialized hardware accelerators, optimized software, and communication protocols are explored. Challenges including resource limitations, security, model management, power consumption, and connectivity are critically assessed. Emerging opportunities in neuromorphic hardware, continual learning algorithms, edge-cloud collaboration, and trustworthiness integration are highlighted, providing a comprehensive framework for researchers and practitioners.

---

### AFFORD2ACT: Affordance-Guided Automatic Keypoint Selection for Generalizable and Lightweight Robotic Manipulation

Authors: Anukriti Singh, Kasra Torshizi, Khuzema Habib, Kelin Yu, Ruohan Gao, Pratap Tokekar
Published: 2025-10-01
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2510.01433v2

Vision-based robot learning often relies on dense image or point-cloud inputs, which are computationally heavy and entangle irrelevant background features. Existing keypoint-based approaches can focus on manipulation-centric features and be lightweight, but either depend on manual heuristics or task-coupled selection, limiting scalability and semantic understanding. To address this, we propose AFFORD2ACT, an affordance-guided framework that distills a minimal set of semantic 2D keypoints from a text prompt and a single image. AFFORD2ACT follows a three-stage pipeline: affordance filtering, category-level keypoint construction, and transformer-based policy learning with embedded gating to reason about the most relevant keypoints, yielding a compact 38-dimensional state policy that can be trained in 15 minutes, which performs well in real-time without proprioception or dense representations. Across diverse real-world manipulation tasks, AFFORD2ACT consistently improves data efficiency, achieving an 82% success rate on unseen objects, novel categories, backgrounds, and distractors.

---