# Research Papers — 2026-08 · bundle page 41 of 240 > Full-text markdown of papers #2001–2050 of 11995 published in 2026-08, newest first. AI-readable mirrors of the full 11995-paper 2026-08 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-08/page/41 [← Previous bundle](https://aipod.dev/papers/archive/2026-08/page/40.md) [Next bundle →](https://aipod.dev/papers/archive/2026-08/page/42.md) ### The Limits of Speculation: Bounding Speculative Decoding in Mixture-of-Experts Authors: Aidar Amankulov, Denis Mamatin Published: 2026-08-26 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.22156v1 Speculative decoding in Mixture-of-Experts (MoE) models faces the problem of unstable verification cost caused by input-dependent expert loading. To study the physics of this process, we formulate speculation-budget selection as an offline Stochastic Shortest Path (SSP) problem over reference sequences and build a diagnostic Oracle that uses counterfactual simulation to account for MoE verification cost. A detailed analysis of the Oracle's decisions on the Qwen3-Coder and EAGLE-3 pairing, in the space of marginal deltas (Delta Space), shows that rejected candidates form a strict linear boundary. This result demonstrates that a complex global optimization is locally governed by a necessary condition balancing marginal cost against expected progress ($\frac{Δ\mathbb{E}(Cost)}{Δ\mathbb{E}(a)}$), providing a rigorous mathematical reference point for designing future adaptive online heuristics. --- ### From Latent Biomarkers to Clinical Rules: Embedding-Guided Rule Mining and Attribution-Based Translation for Interpretable Tabular Learning Authors: Majid Lotfian Delouee, Hamed Ayoobi, Sjors G. J. G. In 't Veld, Martijn C. Schut Published: 2026-08-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.22155v1 Clinical decision support tools are most useful when accurate predictions are accompanied by understandable explanations. Rule-based models provide transparency, but rules derived directly from raw clinical measurements may miss patterns arising from interactions between multiple variables. We present a four-step pipeline that mines decision rules in the latent space of an FT-Transformer and translates them back into measurable clinical features. Embedding dimensions that consistently separate patient groups are treated as latent biomarkers, rules are mined using small decision trees, and selected rules are translated using gradient-input saliency and CLS attention attribution. We evaluate the framework on six public clinical and population health datasets at four embedding dimensions. Translated rules outperformed raw-feature rules in five of six datasets, with mean AUROC gains ranging from 0.04 to 0.23. On the heart disease dataset, embedding-space rules reached 0.98 AUROC, but translation reduced this to 0.72, showing that high-performing latent rules cannot always be represented by simple raw-feature conditions. These results show that latent-space rule discovery can uncover predictive patterns while translating them into clinically measurable features that can be evaluated by clinicians. --- ### A Comparative Framework for Evaluating Foundation Models on Tabular Data: A Case Study in Healthcare Authors: Majid Lotfian Delouee, Sjors G. J. G. In 't Veld, Martijn C. Schut Published: 2026-08-26 Categories: cs.LG, cs.CY arXiv: https://arxiv.org/abs/2609.22154v1 Tabular data is the most common format in clinical practice, encompassing laboratory results, medication records, diagnostic codes, and patient demographics. As foundation models for tabular data have grown in number and variety, a practical question has become harder to answer: which model should a clinician or data scientist actually choose for a given task, and why? Existing surveys catalogue what these models can do, but they stop short of providing a structured way to compare them against the specific demands of a real application. We introduce \system{}, a comparative evaluation framework that scores and ranks tabular foundation models (TFMs) across six clinically meaningful dimensions: how well a model generalizes to new datasets, how effectively it protects patient privacy, how much data it needs to perform well, how it scales with growing datasets and feature spaces, how interpretable its predictions are to clinicians, and how fairly it performs across patient subgroups. Each dimension is broken down into measurable sub-components, and groups of sub-components can optionally be combined into supplementary compound scores, called super-metrics, that provide a diagnostic view of how a model performs across several dimensions simultaneously. To show how the framework works in practice, we apply it to two healthcare use cases, screening for iron deficiency and predicting heart failure, demonstrating how the same set of metrics leads to different model rankings depending on what matters most in each clinical context. We also provide a taxonomy of 45 TFMs organized by their underlying architecture, which serves as a reference for researchers and practitioners looking to navigate this rapidly expanding field. --- ### SafeTune: A Unified Faithful Library for Auditing and Repairing Safety Drift in Fine-Tuned LLMs Authors: Pratinav Seth, Saisab Sadhu, Anshul Kaushal, Vinay Kumar Sankarapu Published: 2026-08-26 Categories: cs.LG, cs.SE arXiv: https://arxiv.org/abs/2609.22153v1 Methods for addressing safety drift in fine-tuned Large Language Models (LLMs) are scattered across incompatible implementations, lifecycle stages, and evaluation protocols, making them difficult to adopt and compare. We introduce SafeTune, a source-available library that unifies four intervention paradigms: post-hoc weight recovery, safety-constrained fine-tuning, gradient-based unlearning, and inference-time steering, alongside shared interpretability, evaluation, and deployment utilities. SafeTune provides a consistent configuration-driven workflow while preserving the distinct inputs and intervention points each paradigm requires. Its modular registry supports new methods, benchmarks, judges, models, and fine-tuning domains without redesigning the surrounding pipeline. We demonstrate SafeTune through controlled comparisons and finance and medical deployment case studies, showing how it characterizes safety drift, evaluates feasible interventions on common refusal-behavior and capability evaluations, and supports calibrated or layered mitigation. --- ### Is Imagination Derived from Hallucination? A Cross-Taxonomy Evaluation of Imagination and Hallucination in Large Language Models Authors: Zixuan Tang, Hongzong Li, Shuxin Zhuang, Dapeng Wu, Zi Liang Published: 2026-08-26 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.22152v1 Imagination performs as a high-level function of large language models (LLMs) which determines the potential of how an LLM creates unseen or creative content. While existing works have built a rich family of creativity benchmarks for this ability, they only measure how far an output departs from common answers and never check whether the departure is licensed by the prompt. Moreover, hallucination, the closest neighbor of imagination, is always measured in a separate pipeline on different generations, so the influential claim that imagination and hallucination stem from the same generative mechanism has never been directly testable. In this paper, we propose Whiteboard, the first LLM imagination evaluation benchmark. Its design follows the authoritative cognitive instruments developed to measure human imagination: seven mechanism-grounded imagination subtypes are adapted from classic paradigms, then crossed with ten support-boundary hallucination subtypes and scored jointly on the same generation. Different from previous creativity or hallucination benchmarks, Whiteboard gates every imagination score with an explicit support check and computes both axes deterministically through an auditable atom matrix, with no LLM judge on the primary path. The full Whiteboard item bank contains 1,660 prompts; on its shared 80-item anchor set, we evaluate 79 state-of-the-art LLMs and validate the instrument against 13,280 human judgments. Additionally, we further explore whether imagination derives from the same generative tendency as hallucination and what key factors shape it. Our analysis indicates a counterintuitive correlation between hallucination and imagination: Most of the subtype couplings are negative, every one of the anchor items reproduces the negative coupling on its own. --- ### Do Language Models Know Their Own Constraints? Authors: Arin Agarwal Published: 2026-08-26 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.22151v1 We ask whether behavioral constraints acquired through post training remain explicitly reportable. Using constrained recipe generation as a testbed, five banned ingredients enforced via LoRA fine tuning of Llama 3.1 8B Instruct we compare supervised fine tuning (SFT) and Group Relative Policy Optimization (GRPO) against an untrained baseline on a four tier Constraint Awareness Benchmark. Averaged over three seeds, both methods raise behavioral compliance from 4% to about 90% while reducing explicit constraint reporting below the untrained model (0.48/5 to 0.16/5 for SFT, 0.07/5 for GRPO) and eroding retained third person knowledge (93% to 36% for SFT, 14% for GRPO; p less than 0.01 between methods). Contrary to our initial hypothesis, the reward based signal is the more destructive of the two: a reward that penalizes banned ingredient tokens regardless of framing learns a context independent suppression rather than a self directed constraint. A context conditioned reward designed to teach the self to other distinction fails, collapsing toward inclusion in both framings. Probing prompt time hidden states recovers per ingredient avoidance at 83.8% (layer 24 MLP), but only 6.4 points above a per ingredient base rate predictor (77.4%), and the model's own verbal self report is more accurate still (87.8%). A positive control adding explicit self description examples does not restore reporting. The failure is therefore specific to enumerating constraints on request, not a general loss of access to them. --- ### ECP-Bench: Benchmarking and Learning Entertainment Content Promotion with Foundation Models Authors: Hyomin Kim, Bowen Chen, Jin Huang, Zhao Wang, Qiaozhu Mei, Shingo Takamatsu Published: 2026-08-26 Categories: cs.IR, cs.LG arXiv: https://arxiv.org/abs/2609.22150v1 Content promotion spans a broad set of skills, from understanding content to forecasting its market reception. However, LLMs' ability to support such promotion decisions remains underexplored. Existing studies are often limited to a single task (e.g., popularity prediction) or a small set of tasks within a single domain (e.g., movies). As a result, there is a lack of understanding of LLMs' abilities in the full promotion process and how these abilities generalize across different tasks and domains. In this work, we introduce ECP-Bench, a benchmark containing 1.9M movie, game, and music items and 423,451 questions across 33 tasks in five content-promotion skill families. Our evaluation shows that frontier models achieve only 51.9\% overall accuracy and lose much of their advantage on post-cutoff content, with drops of up to 19.1 percentage points. In contrast, open-weight models fine-tuned on ECP-Bench achieve up to 60.3\%, remain substantially more stable across the knowledge cutoff, generalize to unseen content and tasks, and exhibit meaningful cross-domain generalization. --- ### Beyond the Stitching Assumption: A Unified Framework for Multimodal Synthetic Data Evaluation via Semantic Quantization Authors: Yefeng Yuan, Zhan Shi, Liang Cheng, Yuhong Liu Published: 2026-08-26 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.22149v1 Multimodal synthetic datasets combine structured attributes with free text, but are often evaluated separately. Such metrics can remain high after tabular--text pairings are disrupted. We present a projection-based evaluator for tabular--text synthetic data. A fixed sentence encoder maps text to embeddings, \(k\)-means converts them to cluster states, and tabular variables are represented as categorical or quantile-binned states. Real and synthetic contingency tables are compared using Jensen--Shannon divergence (JSD), normalized mutual information (NMI), conditional JSD (cJSD), and joint-state entropy. We also report text-to-attribute (T2A) utility and a holdout-calibrated proximity flag rate (PFR) as a representation-level diagnostic. A text-permutation control preserves both marginal distributions while disrupting their pairing. Experiments on Amazon Reviews, Kiva Loans, and the Employment Scam Aegean Dataset show that modality-specific scores remain high under this control. The projection diagnostics detect disruption when real projected dependence exceeds a permutation baseline, but are less informative for weak or sparse projections. Some conditioned LLM baselines also exhibit stronger measured dependence than the corresponding real-data projections. These results support explicit cross-modal evaluation with permutation baselines and coverage reporting. --- ### GRRR: The Geometry of Reshaping, Rotation, and Routing in Decoder LLM post-training Authors: Jianing Qi, Hao Tang, Zhigang Zhu Published: 2026-08-26 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2609.22146v1 We study how post-training changes the weights of Large Language Models (LLMs) relative to their pretrained weights. Across 12 post-training chains with supervised fine-tuning (SFT) and reinforcement learning (RL), we express each weight update in the pretrained matrix's singular value decomposition (SVD) frame. This decomposition separates the changes of three geometrically distinct components: diagonal values, which reshapes singular values; off-diagonal values, which rotates the coupling between pretrained input and output directions; and null-space values, which routes outside the matrix's original nonzero SVD core. On a math evaluation suite, we find that removing the diagonal component usually preserves most of the gains from post-training. These results suggest that post-training gains are carried primarily by reconfiguring and extending pretrained pathways rather than by substantially changing singular values of pre-trained models. --- ### Weak Ties, Strong Signals: Efficient Training Data Detection in Diffusion LLMs via Independent Token Sampling Authors: Hongyao Yu, Tianqu Zhuang, Ziyuan Xu, Hao Fang, Jiaxin Hong, Bin Chen, Shu-Tao Xia Published: 2026-08-26 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2609.22145v1 Diffusion large language models (dLLMs) offer a compelling alternative to autoregressive models, yet they may expose sensitive training data during denoising. Detecting such usage is challenging because dLLMs lack the efficient one-pass probability decomposition of causal architectures. Existing methods rely on random masking to obtain tractable token-wise detection signals under limited query budgets, but fail to control dependencies among masked tokens. We demonstrate that this token-wise approximation introduces a non-negative structural estimation error, which is theoretically characterized by the cumulative conditional mutual information (CMI) among masked tokens and can obscure subtle memorization signals. This insight suggests that reliable detection requires masked token sets with weak internal dependency. To avoid the prohibitive cost of directly estimating CMI over token combinations, we propose \textit{Independent Token Sampling} (ITS), a query-efficient framework that uses an attention-derived pairwise dependency proxy to approximate the CMI-aware selection criterion. ITS further incorporates a diversity-promoting strategy to improve token coverage across sampling rounds, yielding aggregated token-wise signals that are less affected by dependency-induced approximation error. Experiments on multiple datasets show that ITS consistently outperforms state-of-the-art baselines across different models and datasets, achieving an AUC improvement of 0.18 on the ArXiv dataset while maintaining strong performance under limited query budgets. The code is available at https://github.com/Chrisqcwx/DLLM-MIA . --- ### Multilingual Safety Signals Are Multi-Layered: Filtering Safety-Degrading Data for Safer LLMs Authors: Jiakun Li, Guowei Song, Sijia Li, Xingwei He, Hongzheng Chai, Yuan Yuan Published: 2026-08-26 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.22144v1 Preserving safety alignment during large language models fine-tuning is critical, however, recent studies have demonstrated that even benign fine-tuning data may contain safety-degrading samples that silently undermine safety alignment. Existing approaches typically identify such samples using representations from a single safety-sensitive layer. While this assumption has shown effectiveness in monolingual settings, its validity for multilingual models remains unclear due to potential cross-lingual differences in representation patterns. Through a cross-lingual analysis, we show that sensitive layers are only partially shared across languages, with safety-relevant signals often distributed across multiple layers. Motivated by these observations, we propose MMSAFE, a multi-layer framework for multilingual safety-degrading data identification that captures both shared and language-specific safety signals. Extensive experiments across multiple models, languages, and safety benchmarks demonstrate that MMSAFE reduces the average harmful-response ratio by 60% compared with random filtering and achieves stronger average performance than the strongest single-layer baseline, demonstrating the effectiveness of multi-layer modeling for robust multilingual safety alignment. --- ### Impact of Multiple Non-Invasive Biosignals on Cardiovascular Biomarker Estimation via Simulation-Based Inference Authors: Shusaku Maeda, Masahiro Nakano, Tomoharu Iwata, Kenji Komiya, Ryo Nishikimi, Kunio Kashino Published: 2026-08-26 Categories: physics.med-ph, cs.LG, stat.AP arXiv: https://arxiv.org/abs/2609.11969v1 As the population ages, the number of patients with cardiovascular diseases continues to increase, highlighting the need for early detection before progression to severe and irreversible functional decline. Consequently, estimating cardiovascular biomarkers from non-invasive biosignals, such as photoplethysmography (PPG) and arterial pressure wave (APW) signals, has attracted increasing attention. These signals can be measured using wearable and cuff-type devices. Previous studies have used PPG and APW signals to estimate cardiovascular biomarkers. However, these signals exhibit strong similarities in both the temporal and frequency domains and primarily reflect peripheral and arterial pulse waveforms. Therefore, they may provide limited information about cardiac mechanical function. In contrast, the quantitative impact of additional biosignals, such as ballistocardiography (BCG), which reflect the body's minute mechanical responses to cardiac ejection, remains unclear. In this study, we generated synthetic PPG, APW, and BCG signals from a unified whole-body cardiovascular circulation model and evaluated the complementary contribution of BCG to probabilistic cardiovascular biomarker estimation. We estimated posterior distributions of cardiovascular biomarkers using neural posterior estimation and simulation-based inference. The results showed that adding BCG signals significantly improved estimation performance. Furthermore, even in ill-posed cases where PPG and APW alone produced multimodal posterior distributions, adding BCG yielded unimodal posterior distributions. These findings provide fundamental insights into signal selection for estimating cardiovascular dynamics. --- ### Optimizing Train Driving to Minimize the Electricity Cost of an Entire Railway Traffic Mesh using Evolutionary Algorithms Authors: Eduardo Pilo de la Fuente, Sudip K. Mazumder, Maria Antonia Simon Published: 2026-08-26 Categories: cs.NE, physics.app-ph arXiv: https://arxiv.org/abs/2609.05493v1 This paper presents a procedure to optimize the way trains are driven, which pursues, in addition to fulfilling operational constraints such as admissible speeds or journey durations, the minimization of the cost related to supplying electrical energy to the trains, including the cost of the energy consumption and the cost of the power capacity utilization. This procedure combines: (i) a traffic model that merges the energy and power footprint of each rail service part of the traffic mesh, and (ii) an evolutionary computation framework that enables searching for the optimal way to drive the trains to achieve an optimal traffic mesh. This procedure is applied to a 450 km long section of the high-speed line from Madrid to Barcelona (Spain). --- ### Towards Scaling Reinforcement Learning to Massive Populations: Learning Mean-Field Representations Authors: Aditya Makkar, Benjamin Unger, Jeongyeol Kwon, Mathieu Laurière, Eugene Vinitsky, Yonathan Efroni Published: 2026-08-26 Categories: cs.MA, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.02928v1 Modern multi-agent systems are increasingly deployed at scale over large populations of agents in settings such as ad-auctions, traffic routing, and recommendation systems. The dominant approach in such settings is to optimize each agent's policy independently, treating the other agents as part of a fixed single-agent environment rather than modeling the population dynamics. In many large-population systems, the dynamics depend on an aggregate summary of the population rather than the identity of any individual. Mean-field RL exploits such structure, providing a principled framework that models each agent's environment as an explicit function of the population distribution. However, in large state-action spaces or high-dimensional control problems, modeling the population distribution is itself intractable. How can we design a scalable framework for high-dimensional control problems with large populations? This work explores this question from the perspective of representation learning. We introduce a mean-field RL framework in which the rewards and transition dynamics depend on the population only through an unknown low-dimensional aggregate statistic. We then study this framework in the offline setting and design a provable approach that learns a near-optimal policy by learning a low-dimensional representation. Motivated by real-life supply-chain optimization problems, we design a one-step routing game to test the hypothesis that learning a low-dimensional population representation improves reward prediction and Nash gap estimation relative to baselines that don't exploit this structure. We show that under a fixed neural-network parameter count and optimization budget, learning a low-dimensional population representation improves reward prediction and the equilibrium quality of the resulting policies. --- ### ES-AHD: An Evolution Strategy Framework for Automatic Heuristic Design Authors: Yutao Lai, Kezhao Lai, Hai-Lin Liu, Yuping Wang, Ping Guo Published: 2026-08-26 Categories: cs.NE, cs.LG arXiv: https://arxiv.org/abs/2609.00023v1 In this paper, we introduce ES-AHD, a novel framework that fundamentally integrates Evolution Strategy (ES) into Large Language Model (LLM)-driven Automatic Heuristic Design (AHD). Existing evolutionary approaches predominantly rely on random, individual-level mutation, leading to blind search and an imbalance between exploration and exploitation. To address these issues, ES-AHD introduces two core mechanisms. First, Semantic Recombination via LLMs discards traditional point-to-point reproduction. By leveraging the LLM's contextual reasoning to explicitly extract core insights from top-performing individuals, the algorithm establishes a promising semantic search direction. This transforms random code mutation into targeted, center-guided sampling inspired by ES. Second, Stochastic Covariance Adaptation via Temperature Sampling dynamically addresses the exploration-exploitation dilemma. By mapping the covariance matrix in ES to the LLM's sampling temperature, the framework employs a stochastic random walk mechanism with momentum. This approach primarily shrinks the search radius for micro-level code refinement, while retaining the critical ability to occasionally sample higher temperatures to escape semantic local optima. Ultimately, ES-AHD provides a highly directional, robust, and efficient search paradigm, significantly accelerating the generation of high-quality heuristic algorithms. The source code is available at: https://github.com/Mriya0306/ES-AHD. --- ### Defending Wearable VLMs Against Private Attribute Inference Authors: Zhimin Li, Pan Wang, Jingxian Chen, Yuantao Tang, Anthony Chen, Qian Lou, Jingtong Hu Published: 2026-08-26 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2608.28691v1 Wearable VLM pipelines promise continuous multimodal assistance from egocentric visual capture: a user asks a task-driven question about the surrounding scene, and the system uses compact visual tokens to support language reasoning. The challenge motivating this work is that the same egocentric evidence needed for useful assistance can also reveal private attributes about the wearer or nearby bystanders. We investigate this as a joint privacy-utility problem for split VLM inference, where visual encoding occurs within a trusted device boundary but intermediate visual tokens may be transmitted to downstream reasoning components. This exposes an understudied leakage surface: even when final textual responses are benign, external attackers or untrusted downstream components can recover private attributes from transmitted visual tokens. To evaluate this tension, we construct a paired privacy-utility benchmark with 3,221 image-question records, each paired with a utility question and privacy labels covering location, income, sex, and interests. We further propose Token-Guided Attribute Privacy (TGAP), a pre-LLM token disentangler that learns a residual transformation of visual tokens before they leave the trusted boundary. TGAP combines utility preservation, identity regularization, semantic privacy suppression, and image-driven representation suppression, avoiding the utility loss caused by coarse hard or attention masking. On the benchmark used for source-model evaluation, TGAP reduces privacy accuracy from 56.7\% to 7.4\%, a 49.3\% absolute drop, while maintaining relaxed utility at 74.4\%. These results suggest that securing the compact token interface is a practical path toward privacy-preserving wearable multimodal AI. --- ### Projection-Aware End-to-End Learned Video Compression for 360-Degree Video Authors: Niloofar Maani Published: 2026-08-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2608.28689v2 360-degree video supports immersive applications such as virtual reality, autonomous driving, and education. Because spherical content cannot be processed directly by conventional video codecs, it must first be mapped to a two-dimensional projection. Projection choice affects spatial continuity, sampling uniformity, motion estimation, and compression efficiency. This thesis investigates how projection format influences end-to-end neural compression of 360-degree video. Seven formats supported by JVET 360Lib are evaluated using the scale-space flow model, JVET test sequences, and common test conditions. Each sequence is converted from its source equirectangular projection to a coding projection, compressed at multiple rate points, reconstructed, and converted back. Performance is assessed using PSNR, spherical PSNR, weighted spherical PSNR, and Bjøntegaard delta rate. A differentiable pipeline combining projection conversion, neural compression, and inverse projection is also compared with 360Lib. Results show that equirectangular and padded equirectangular projections provide the highest compression efficiency with the scale-space flow model, while cubemap-based and rhombic dodecahedron projections are less effective. This differs from the conventional HM-16.16 codec, for which cubemap-based formats, particularly equi-angular and adjusted cubemap projections, outperform equirectangular formats. Neural models based on optical flow benefit from the spatial continuity of single-face projections, whereas block-based hybrid codecs better accommodate multi-face layouts. These findings show that projection efficiency is codec-dependent and provide guidance for selecting projections for learning-based 360-degree video compression. --- ### FLM: Frequency-Aware Language Models for Generative Image Compression Authors: Jiarun Chen, Kejun Wu, Li Li, Chengtao Cai, Zhengguo Li, Chia-Wen Lin Published: 2026-08-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2608.28687v1 Generative models have significantly improved the performance ceiling of image lossy compression at low bitrates by exploiting learned priors. However, the generated textures and semantic details may deviate from the source content, thereby affecting the fidelity of image reconstruction. To solve these challenges, we propose FLM, a frequency-aware language model that improves compression efficiency through frequency-domain probabilistic modeling while retaining deterministic reconstruction. At the encoder, the input image is transformed into quantized DCT coefficients, which are organized into discrete sequences using macroblock-based coefficient tokenization. FLM then performs next-coefficient prediction to autoregressively estimate token-wise conditional probability distributions for arithmetic coding, thereby generating a compact bitstream. At the decoder, the LLM and arithmetic decoder jointly recover the frequency-domain data, followed by inverse transformations for image reconstruction. A task-specific frequency-domain dataset and a two-stage fine-tuning strategy are further developed to enable the model to operate across multiple bitrate settings. FLM is a versatile compressor that is compatible with both lossy compression and lossless JPEG recompression frameworks. Experiments show that FLM exceeds conventional and generative lossy compression methods in rate-distortion performance. FLM achieves BD-PSNR gains of 3.30 dB, 3.83 dB, and 3.80 dB than JPEG baseline on Kodak, Tecnick, and CLIC2020, respectively. Better qualitative quality of FLM can be achieved in improving semantically high fidelity and suppressing blocking artifacts. FLM is also validated to be applicable to the lossless recompression task with competitive performance. --- ### Data Diversity, Not Frequency Invariance: A Controlled and Self-Audited Study of Compression-Robust Deepfake Detection Authors: Abbas Aliyev, Samir Rustamov Published: 2026-08-26 Categories: cs.CV, cs.LG, cs.MM arXiv: https://arxiv.org/abs/2608.28685v1 Frequency features and compression-invariant representation learning are widely assumed to be key to deepfake detection that survives video compression. We test this with CAFRL - block-DCT and FFT-phase streams, compression-level-conditioned band attention, and adversarial (gradient-reversal) compression invariance - and report a controlled negative. Under a pre-registered protocol with capacity- and augmentation-matched controls, a plain EfficientNet-B0 on multi-quality data beat CAFRL as specified at every compression level on the FaceForensics++ test split, by 3.66 AUC points at CRF 40 (paired, single seed). A self-audit of our own negative found four defects biased against the frequency hypothesis, and pre-specified re-tests repairing all four showed the deficit to be a recipe artifact, not an architecture failure: the baseline recipe recovered 3.96 points over the matching shipped-recipe variant. The frequency path made no detectable difference: discriminative alone (standalone validation AUC 0.91-0.98 late in training) but of no marginal value under this fusion, at two feature widths of one 4.0 M trunk, every seed-pooled interval for the intra-dataset compression contrasts including zero; on the single held-out manipulation tested, the fair variants sat below the plain backbone. The adversarial branch, as specified, added nothing and degraded its own conditioning estimator; at the fair recipe it is untested. Robustness under single-pass H.264 re-encoding came instead from data diversity: real constant-rate-factor variants beat synthetic JPEG augmentation by 7.3 points (single runs, non-overlapping intervals). The evidence is FaceForensics++-family, GAN-era and single-codec. Match controls on training recipe as well as capacity, and buy compression robustness with codec diversity before architecture. --- ### Distributed Semantic Segmentation With Improved Rate-Distortion Trade-Off Authors: Danish Nazir, Timo Bartels, Thorsten Bagdonat, Tim Fingscheidt Published: 2026-08-26 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2608.28684v1 Distributed deep neural networks (DNNs) for dense perception tasks such as semantic segmentation execute an encoder DNN on edge devices, and a decoder DNN typically on a large-scale cloud platform with a particular constraint on transmission bitrate. Recent works employ source codecs to enable bitrate-efficient transmission between the edge device and the cloud. However, as these approaches are typically bound to a particular type of source codec and alternative network architectures are often not explored, this results in a suboptimal rate-distortion (RD) trade-off in the low-bitrate regime. In this work, we propose two novel source codecs that \textit{enable extremely low bitrates, while improving RD performance}. We demonstrate the effectiveness of our proposed source codecs by achieving state-of-the-art performance in distributed semantic segmentation at below 0.2 (0.03) bits per pixel, measured using the mean intersection-over-union metric on ADE20K (Cityscapes). --- ### CARD: Calibration via Agreement in Reverse Diffusion for Out-of-Domain MRI Segmentation Authors: Jiaheng Dai, Weidong Guo, Qingbiao Li, Jie Xu, Yi Guo, Yuanyuan Wang, Zeju Li Published: 2026-08-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2608.28681v1 Probability calibration aligns model confidence with predictive accuracy, enabling clinicians to identify unreliable segmentation regions. This alignment breaks down under domain shift, where artifacts and unseen protocols produce confident errors. Existing post-hoc methods adapt the correction at test time, conditioning on predictive entropy, the logit pattern, or augmentation response, but each proxy is read from the terminal prediction, the very quantity that shift corrupts. This motivates reliability evidence beyond the terminal prediction, which categorical diffusion provides in two ways. First, a generative shape prior keeps a capacity-limited reference intact when appearance is corrupted, so its disagreement with the primary segmentor highlights primary-model errors. Second, every reverse step yields a class distribution, separating persistent disagreement from transient discrepancy. Aggregated over the trajectory, this disagreement correlates with Dice at 0.788, against 0.521 for a matched discriminative control. We therefore propose CARD (Calibration via Agreement in Reverse Diffusion), which maps the temporal aggregate of this disagreement to a temperature field applied per pixel across all classes, so that confidence changes while the segmentation does not. Across cardiac, prostate and brain MRI shifts, CARD lowers calibration error in 45 of 49 comparisons against the strongest baseline in each setting. --- ### Multi-Sensor Mapping of Vulnerable Urban Settlements Using SAR, Multispectral, and Hyperspectral Imagery: A Case Study in Córdoba, Argentina Authors: Luigi Russo, Anabella Ferral, Silvia Liberata Ullo, Paolo Gamba Published: 2026-08-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2608.28680v1 Informal settlements represent a major urban challenge in rapidly expanding cities, yet their identification from Earth Observation (EO) data remains difficult because of their heterogeneous appearance and incomplete official inventories. This work presents a multi-sensor deep learning (DL) framework for slum-likelihood mapping in Córdoba, Argentina, integrating high-resolution PlanetScope multispectral (MS) imagery, COSMO-SkyMed (CSK) Synthetic Aperture Radar (SAR) data, and medium-resolution PRISMA hyperspectral (HS) observations. The problem is formulated as a patch-level classification task using the official Registro Nacional de Barrios Populares (ReNaBaP) inventory as reference, and the models are evaluated through four geographically partitioned folds. SAR-only and MS-only baselines, their configurations with PRISMA HS support, and early fusion (EF), middle fusion (MF), and late fusion (LF) strategies are systematically compared. Results show that LF+HS provides the best overall balance between classification performance and spatial selectivity, while PRISMA contributes complementary spectral information alongside the higher-resolution MS and SAR representations. Beyond the standard evaluation against ReNaBaP, an external municipal vulnerability layer is used to interpret detections outside the official polygons, showing that several apparent false positives overlap broader vulnerable urban areas. Thermal analysis further shows that ReNaBaP settlements exhibit significantly higher surface temperatures than their immediate surroundings during a heatwave event, indicating localised surface-heat amplification. Taken together, these results suggest that multi-sensor EO fusion can support both the mapping of ReNaBaP settlements and the interpretation of broader urban vulnerability patterns. --- ### Evaluating Constrained Iterative Refinement for Scalable Vector Graphics Generation with Off-the-Shelf VLMs Authors: Matthew Perlman, James Beetham, Niels Da Vitoria Lobo, Amrit Singh Bedi, Mubarak Shah Published: 2026-08-26 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2608.28678v1 Scalable Vector Graphics (SVGs) power much of the modern visual ecosystem, yet state-of-the-art generative models focus almost entirely on rasterized images. We explore whether inference-time methods can unlock SVG generation capabilities in off-the-shelf vision-language models (VLMs). We systematically evaluate a constrained iterative refinement harness that combines visual feedback, structured editing, and constrained decoding to characterize the capabilities and limitations of current VLMs for SVG generation. Across multiple VLMs and generation settings, we find that constrained decoding improves compilation success rates, while iterative refinement reveals a deficit in visual reasoning and self-correction. Our results highlight both the promise and current limitations of using inference-time methods to adapt general-purpose VLMs for SVG generation. --- ### Can Tainted Pixels Expose Deepfake Videos? Authors: Juan Hu, Shaojing Fan, Sanjay Saha, Marc Herrera, Terence Sim Published: 2026-08-26 Categories: cs.CR, cs.CV arXiv: https://arxiv.org/abs/2608.27492v1 Publicly-acceesible face-manipulation tools have made deepfake creation accessible to non-expert users. Against these, existing defenses are mostly post-hoc, detecting only after forgery has occurred, and operating on still images rather than videos. Research is lacking in i) the proactive protection of published facial videos against black-box manipulation tools, and in (ii) understanding its perceptual effect on human viewers. We introduce TaintedPixels, a proactive video-protection method built around an asymmetric visibility trade-off: the embedded watermark should remain inconspicuous in the published video but become obvious once a downstream tool manipulates the video. TaintedPixels injects structured periodic perturbations into the blue channel of facial regions and refines them under stripe-visibility, color-cast, and video-level LPIPS budgets, with lightweight motion-adaptive deployment. We believe TaintedPixels is the first proactive defense designed specifically against black-box manipulation tools rather than image-level pipelines or specific surrogate generators. Across three publicly available off-the-shelf video manipulation tools and two off-the-shelf detectors, TaintedPixels attains the highest forgery fake rate while keeping perturbations small (LPIPS = 0.0042). Our non-expert human study, conducted on a diverse set of 300 video stimuli spanning different lighting conditions, backgrounds, and skin tones, shows that protected source videos draw a 3.26% suspicion rate, while forgeries from protected sources are identified as fake much more often than forgeries from unprotected sources (90.72% vs. 56.71%). This validates the effectiveness of TaintedPixels. --- ### Mapping Woody Vegetation from Multi-Source Imagery and Prediction Fusion for Enhanced Data Efficiency and Accuracy Authors: Kal Backman, Jared Wood, Adam Roff Published: 2026-08-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2608.26471v1 Tree cover maps are a fundamental remote sensing product, used to derive ecological insights about the landscape and are essential to change detection, vegetation mapping and fire monitoring programs. However, comprehensive tree cover mapping requires reliable and high-quality imagery, free of cloud and weather defects to ensure accurate model outputs. Deep learning approaches can generate high quality maps with minimal human intervention but require large amounts of human annotated data to be successful. In this work we propose a framework consisting of methods that aim to improve the data efficiency and robustness of deep learning models using data fusion techniques to segment woody vegetation defined as vegetation over the height of 2m across the state of New South Wales, Australia. To improve robustness against varying image quality, we propose an image composition method that normalizes the imagery and removes defects, whilst also minimizing the reliance on individual image quality by proposing a prediction fusion method. The two methods resulted in an error reduction of 38.2% and 53.6% respectively compared to single-source imagery. To address deep learning approaches' limitation of requiring large amounts of data, we apply label transfer to multiple sources of imagery as a form of data augmentation to improve data efficiency. Learning from multiple image sources was shown to be the biggest improvement in performance, resulting in an error reduction between 28.1% to 76.2% across the different validation experiments, whilst reducing the standard deviation of performance across image dates by a factor of 13. --- ### Active Curriculum Refinement for Reinforcement Learning Authors: Zhenya Liu, Yuxin Chen Published: 2026-08-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2608.26469v1 In many reinforcement learning (RL) domains, environments are connected by prerequisite relations, such as difficulty-increasing edits or parameter increments, which induce a directed acyclic curriculum graph (DAG). Although this structure is often exploited only implicitly, explicitly modeling it can improve training. We introduce PATH, a curriculum-learning framework that performs active learning over the curriculum graph. PATH first expands coverage by sampling diverse curriculum paths and then reallocates training toward regions that remain unmastered. Experiments across diverse environments show that PATH explicitly leverages the graph structure to achieve strong robustness and generalization. --- ### Compositional Generalization via Structural Identification in a Category-Theoretic Framework Authors: Akihiro Maeda, Thomas Seiller, Yohei Oseki Published: 2026-08-26 Categories: cs.CL, stat.ML arXiv: https://arxiv.org/abs/2608.26465v1 Compositional generalization is usually evaluated through model accuracy. We instead ask which structural or lexical identifications make held-out COGS examples admissible from the structures observed in training. Sentences are represented as functors from syntactic addresses to lexical tokens, and selective collapses induce Kan extensions that propagate observed associations. Across 21 COGS generalization types, admissibility follows distinct identification profiles, while residual failures separate unsupported structural templates. These data-side diagnoses characterize what the training corpus licenses under specified identifications, without training a predictive model. --- ### Diff Mining: Logit Differences Reveal Finetuning Objectives Authors: Greg Kocher, Robert West, Clément Dumas, Julian Minder Published: 2026-08-26 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2608.26462v1 Finetuning has become the gold standard for refining existing behaviors and inducing new ones in language models, yet it often remains unclear exactly which behaviors emerge during this process. As models grow ever more capable, understanding finetuning better becomes increasingly important, particularly since unwanted behaviors may arise during finetuning. In this paper, we introduce Diff Mining, a simple yet effective framework for identifying what a finetuned model has learned by comparing its logits to those of its base model. Diff Mining effectively surfaces salient tokens that are amplified in the finetuned model, serving as a fingerprint of its training -- even on text unrelated to the finetuning domain. Unlike many existing model diffing methods which require model internals, Diff Mining only needs access to output logits and scales to large models. The framework consists of two modular stages: (i) extracting per-context logit differences between the finetuned and base models on a reference corpus, and (ii) aggregating the resulting signals to construct an interpretable token set representing the finetune. For aggregation, we explore both a simple Top-K frequency method and a Non-negative Matrix Factorization (NMF)-based approach for disentangling multiple finetuning objectives into distinct token clusters. Empirically, Diff Mining succeeds across diverse settings: on finetune domain detection, it significantly outperforms state-of-the-art model diffing methods both in identifying relevant tokens and in downstream performance when an interpretability agent is given access to the extracted token set; on models with injected biases, it identifies more than one third of the biases without targeted probing. Overall, our framework shows promise in developing auditing tools to detect finetuning objectives. --- ### Distributed Training using an Intelligent Network Authors: Nihar Shah, Ben Blier Published: 2026-08-26 Categories: cs.LG, cs.DC, cs.NI arXiv: https://arxiv.org/abs/2608.26453v1 Distributed training across a wide area network (WAN) is challenging, as continuous parameter exchange by islands of compute is constrained by limited bandwidth, high latency, and uneven topology. We propose making the network an active participant in training. On the systems side, such networks should leverage (i) multicast technology to replicate outbound traffic and (ii) in-line FPGAs to aggregate inbound traffic, to ease egress and ingress bottlenecks. These technologies are used for training across workers within a data center, but this paper extends them to the WAN. On the algorithms side, we develop an optimization framework that produces rich synchronization schedules (namely, rotating cliques of islands) around the underlying network topology and these technologies, to maximize information exchange. Finally, we illustrate this on a nine-city topology modeled on the DoubleZero network, a live programmable WAN equipped with both technologies, and show how the optimal schedules shift with the network's capabilities. Together, these can narrow the gap to the gold standard of colocated training. --- ### Toward Equitable Low-Carbon Mobility: Fairness-Aware Demand Prediction for Expanding Bike-Sharing Systems Authors: Man Luo, Yixuan Zhao Published: 2026-08-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2608.26451v1 Bike-sharing systems are an important component of low-carbon urban mobility, but continued expansion creates challenges in both cold-start prediction and equitable resource allocation. Newly deployed stations lack historical ridership records, causing a mismatch between training and inference for graph-based models on evolving networks. Historical demand may also encode structural inequalities, as lower ridership in low-income neighborhoods can reflect limited infrastructure access rather than weak latent demand. Models trained directly on such data may therefore reinforce existing mobility disparities. We propose FairGIN, a fairness-aware graph neural network for demand prediction in expanding bike-sharing systems. FairGIN integrates three components. Expansion-Simulated Increment Training stochastically simulates network expansion during training to reduce the cold-start distribution gap. Attention-Based Knowledge Transfer combines station-adaptive temperature scaling with orthogonal embedding alignment to transfer representations from data-rich existing stations to data-sparse new stations. Fairness-Aware Optimization introduces income-stratified regularization and an equity-calibrated deployment score to support more inclusive station placement. Experiments on NYC and Seattle demonstrate that FairGIN achieves state-of-the-art predictive accuracy across diverse expansion scenarios while substantially reducing income-based disparities without compromising overall system efficiency. --- ### Vowel Signs Are Not Letters: A Pre-tokenization Ceiling on Multilingual Tokenizer Fertility Authors: Sajal Regmi, Siddhartha Pudasaini, Chetan Phakami Pun Published: 2026-08-26 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2608.26449v1 Byte-level BPE tokenizers that use the HuggingFace ByteLevel pre-tokenizer inherit GPT-2's word regex, where a word is defined as \p{L}+, one or more Unicode letters. In abugida scripts, vowels are written as combining marks; this pattern therefore splits each word at every vowel sign. Since BPE merges only within a pre-token, those splits persist through training regardless of vocabulary size or corpus composition. We formalise this effect as a training-free lower bound on fertility. Across 26 languages from a parallel corpus, every one of the 17 abugidas is affected, ranging from 1.47x (Tibetan) to 9.02x (Thai), whereas Latin, Cyrillic, Hangul, and Han show exactly 1.00x. For 5 languages, matched tokenizer pairs that differ only in this character class fall within 2.2% of the predicted floor, scoring 4.78 versus 1.58 tokens per word on Nepali. When the Nepali share of the training corpus is swept from 5% to 95%, the broken tokenizer barely shifts at all (1.7%) while the fixed one shifts 33.9%, which separates a structural ceiling from a data shortage without needing to inspect any code. We train three 268M models that differ only in their tokenizer; the fixed variant achieves 4.43% lower held-out Nepali bits per byte at equal compute, and it still leads when given the same bytes with 1.59x the compute. A census of 3,479 HuggingFace repositories finds the letters-only word class present in 63.3% of the most-downloaded text-generation models, accounting for 72.5% of their downloads. GPT-4o's o200k pattern already uses a mark-aware word class, making the repair itself prior art. We quantify its value, show how to recognise its absence from symptoms alone, map which scripts it reaches, measure how widely it is deployed, and release a 65,536-entry Nepali-English tokenizer with a harness that regenerates every number here from public data on a laptop. --- ### Don't Overthink, Don't Underthink: Toward Adaptive Reasoning in Agentic AI Authors: Md Jueal Mia, M. Hadi Amini Published: 2026-08-26 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2608.26442v1 Recent advances in Large Language Models (LLMs) have shown that increased inference-time reasoning can improve performance on complex tasks. However, many existing approaches rely on fixed or preallocated reasoning controls, such as fixed token budgets, pre-execution difficulty estimates, or activation-space interventions, and are often evaluated on standalone reasoning benchmarks rather than full agentic workflows. These assumptions may not hold in agentic AI systems, where reasoning requirements evolve dynamically through planning, tool use, memory retrieval, and agent-to-agent interactions. Consequently, reasoning can become either excessive or insufficient, resulting in unnecessary computation, increased latency, planning drift, excessive tool use, or incomplete solutions. We argue that a major challenge for next-generation agentic AI is not merely how much reasoning a language model should perform, but how it should allocate reasoning according to evolving task demands. We characterize over-reasoning and under-reasoning as recurring failure modes of misallocated reasoning and evaluate them on MATH-500 and the GAIA public validation benchmark. Using tool-decision latency, token consumption, token-limit exhaustion, and answer correctness, our results suggest that cases classified as over-reasoning are associated with higher computational cost without proportional accuracy gains, whereas cases classified as under-reasoning are consistently associated with incorrect or incomplete solutions. These findings motivate future research on adaptive reasoning mechanisms for agentic AI. --- ### Subgraph Filtering for Fair Graph Neural Networks Authors: Haohui Lu, jiyuan Tian, Fangyu Zhou, Shahadat Uddin Published: 2026-08-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2608.26437v1 Graph neural networks (GNNs) can exhibit unfair behavior even when sensitive attributes are excluded from node features, because graph topology and message passing propagate group-correlated signals under sensitive homophily. Existing fairness-aware GNN methods mainly constrain representations or prediction distributions at a global level, without explicitly controlling the local structural pathways through which biased information propagates during aggregation. We propose Subgraph Filtering for Fair Graph Neural Networks (SF-GNN), a lightweight and architecture-agnostic framework that mitigates structural bias at its source. SF-GNN identifies bias-prone edges by combining sensitive homophily with structural propagation amplifiers, including hub participation and triadic closure. It then incorporates stochastic edge filtering into each message-passing step to selectively downweight or remove these edges while preserving the remaining graph structure. Training further incorporates a statistical-parity regularizer with a warm-up schedule to stabilize optimization. Experiments on five benchmark datasets show that SF-GNN achieves consistent fairness improvements while maintaining competitive predictive performance, leading to a better fairness--accuracy trade-off than recent fairness-aware GNN baselines. --- ### NeoTriFuse: Reliability-Aware Multimodal Fusion under Missingness Heterogeneity for Neonatal Mortality Risk Prediction Authors: Jiyuan Tian, Qincheng Shen, Ye Lin, Yu Gao, Haohui Lu Published: 2026-08-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2608.26436v1 Neonatal mortality risk prediction from bedside monitoring data remains challenging due to extreme class imbalance, heterogeneous clinical risk factors, multi-scale temporal dynamics, and substantial missingness. We propose NeoTriFuse, a reliability-aware multimodal fusion framework for missingness-heterogeneous neonatal monitoring data. Unlike conventional multimodal approaches that treat missingness primarily as a preprocessing issue, NeoTriFuse models missingness as an explicit reliability signal that dynamically modulates modality contributions during fusion. The framework integrates static perinatal variables, local-global temporal encoders, and patient-level statistical summaries through reliability-guided gating mechanisms, while jointly optimizing mortality prediction and an auxiliary length-of-stay objective. NeoTriFuse achieves competitive performance, with an F1 score of 0.6736 +/- 0.0216 and an AUROC of 0.9454 +/- 0.0056. Ablation studies indicate that the local-global temporal architecture and patient-level summary branch contribute most substantially to predictive performance, while reliability-aware gating provides additional improvements on threshold-dependent metrics under heterogeneous observation completeness. Sensitivity analyses further suggest stable performance across nearby hyperparameter settings. Overall, the findings support reliability-aware multimodal fusion as a practical approach for neonatal mortality prediction under realistic clinical missingness conditions. --- ### AfriSwitch: A Benchmark for In-the-Wild African Code-Switched Speech Recognition Authors: Gabrial Zencha Ashungafac, Busayo Awobade, Tobi Olatunji Published: 2026-08-26 Categories: cs.CL arXiv: https://arxiv.org/abs/2608.26434v1 Code-switching is pervasive in bilingual African conversation, yet most ASR systems assume monolingual input and are evaluated on curated monolingual benchmarks. We present AfriSwitch, a 61.36-hour human-transcribed benchmark of in-the-wild code-switched speech spanning 16 African languages and language varieties, released with switch-level English span tags, perutterance Code-Mixing Index (CMI), and switch-point counts. Corpus statistics show that mixing behaviour varies widely across African languages along two largely independent axes: how often speakers alternate, and how balanced the mixture is. No single scalar captures how code-switched a language is. Benchmarking five open and commercial multilingual ASR systems zero-shot yields word error rates far above published monolingual figures for the same languages, with the best system averaging 35.93% WER and no system falling below 24% on any language. Africa-targeted training, not model scale or nominal language coverage, best predicts performance. --- ### FedCMAPSS: A Benchmark for Federated Learning in Remaining Useful Life Estimation Authors: Amelia Sorrenti, Matteo Pennisi, Concetto Spampinato, Simone Palazzo Published: 2026-08-26 Categories: cs.LG arXiv: https://arxiv.org/abs/2608.26433v1 Data-driven prognostics and health management has emerged as a key enabler for Industry 4.0, yet the development of robust remaining useful life (RUL) estimation models is often limited by the scarcity of run-to-failure data. While federated learning offers a promising paradigm to collaboratively train predictive models without sharing sensor data, research efforts have operated so far in the absence of a common evaluation framework. To address this gap, this paper introduces FedCMAPSS, a benchmark for federated RUL estimation based on the commonly-used NASA C-MAPSS dataset. We define a set of five standardized tasks designed to simulate real-world industrial challenges, ranging from ideal IID settings to extreme statistical heterogeneity, and conduct a systematic evaluation of state-of-the-art federated optimization algorithms across multiple neural architectures. By establishing reproducible baselines and making the source code and data splits publicly available, this work aims to provide a standard foundation for developing and comparing federated predictive maintenance solutions. --- ### SpeechGym: An Audio-Native Gym for Training Voice Agents via Reinforcement Learning Authors: Jiajun Fan, Jingyuan Li, Prashanth Gurunath Shivakumar, Jia-Hong Huang, Qi Luo, M. Maruf, Ivan Bulyko, Ge Liu, Roger Ren Published: 2026-08-26 Categories: cs.SD, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2608.26432v1 Voice agents must call tools and hold multi-turn dialogue entirely through speech, yet the dominant paradigm trains them in text. Existing frameworks either cascade TTS and ASR around a proprietary voice API, where gradients cannot flow and per-call cost makes on-policy reinforcement learning prohibitive, or stay in text: they measure voice agents but cannot improve them. We present SpeechGym, an audio-native agentic environment in which two omni-modal models converse in native audio, with no external ASR or TTS and no API boundary, over the unmodified tasks, tools and success check of an established text agentic benchmark, so that the interaction modality is the only variable and the loop stays local and trainable end to end. Audio agentic capability does not follow from audio understanding. The failures speech introduces are perceptual rather than reasoning deficits: the agent picks the right tool and the right argument slot but fills it with a value misheard from the waveform, and that single error cascades into a failed call, a retry of the same call, and a wasted step budget. A second failure is behavioural: under an insistent caller the agent performs an unauthorised write and ends the episode believing it helped. Both are trainable, because the environment labels them for free: a call with a misheard argument fails against the database while a correct one succeeds. The obstacle is sparsity, not signal. Outcome-only GRPO is gradient-starved here, since almost every rollout group fails identically, while a per-turn process reward crediting each successful tool call restores variance to nearly every group. Trained this way, the agent transfers with no further tuning to an independently implemented voice benchmark, more than doubling task success and carrying an open-weights model from last place to second on that leaderboard, while using fewer turns and tokens than before training. --- ### Fine-Tuning of Transformer models with Frames Authors: Harshavardhan Adepu, Li Zhang, Sanjiv Kumar, Vikas Singh Published: 2026-08-26 Categories: cs.AI arXiv: https://arxiv.org/abs/2608.26430v1 Parameter-Efficient Fine-Tuning (PEFT) strategies such as Low-Rank Adaptation (LoRA) are effective solutions for fine-tuning large-scale pre-trained models; however, their memory requirements scale with the size of the model, $\mathcal{O}(dr)$, where $d$ is the model's hidden dimension and $r$ is the rank. Our proposal, FrameFT, models the parameter update $ΔW$ with a sparse coefficient matrix in a Fusion Frame basis. Fusion Frames can be generated algorithmically and shared across model layers, enabling very efficient updates. Only the sparse coefficients of the basis expansion are stored/optimized, reducing the memory footprint. The sparse structure of the coefficient matrix in FrameFT and the sparsity in the Fusion Frames give large compute benefits, and our analysis provides formal convergence results. We evaluate the idea across a suite of supervised fine-tuning benchmarks, focusing on language tasks, but also report application to vision models. Our experiments show that FrameFT achieves performance on par with/exceeding state-of-the-art PEFT techniques, but needs far fewer trainable parameters. --- ### The Latent Diagnostic Taxonomy: A Framework for Constructing Classifiers and Diagnosing Their Decisions, Applied to Prompt Injection Detection Authors: Jaturong Kongmanee, Smile Thanapattheerakul Published: 2026-08-26 Categories: cs.LG, cs.AI, cs.CL, cs.CR arXiv: https://arxiv.org/abs/2608.26423v1 This paper proposes a framework for constructing a classifier as a safeguard layer, and for developing a complementary diagnostic that identifies which of the classifier's confident decisions can be trusted. This framework, the Latent Diagnostic Taxonomy, consists of (i) constructing a dimensionality-optimized classifier, in which the embedding dimensionality is empirically selected via cross-validated performance rather than fixed a priori, (ii) locating a relatively small set of latent support vectors (~ 29% of total training examples) representing influential prompts for identifying tokens that alter the classifier's predicted labels, and (iii) utilizing such tokens and their associated attack magnitudes for constructing a diagnostic taxonomy. This diagnostic taxonomy provides an end-to-end guideline for flagging prompts that require different treatments: rely Safely on the classifier's decision; flag Heuristic Bias and Heuristic Override cases; route Insufficient Context cases for further human/safety review. Applying the framework to a classifier trained on a public prompt injection dataset, we find that a substantial fraction of its confident decisions (~ 77%) are not robust to removing a single token, and that this brittleness separates into two distinct failure patterns: a confidence calibration failure and a genuinely exploitable shortcut. For each zone of the taxonomy, we also recommend strategies for remediating diagnosed prompts. We illustrate the framework as a series of steps, demonstrating how each step operates. --- ### Interpreting Latent Protein Language Model Features with Geometric Annotations Authors: Siddharth Setlur, Djordje Mihajlovic, Darrick Lee Published: 2026-08-26 Categories: q-bio.QM, cs.LG arXiv: https://arxiv.org/abs/2608.26419v1 Protein language models (pLMs) encode information about protein sequences which enable downstream tasks such as structure prediction, but their internal representations are not well understood. Sparse autoencoders (SAEs) provide a promising tool to disentangle latent pLM representations into interpretable features, but existing annotation pipelines largely rely on protein-level annotations derived from database labels and LLM annotations of top activating sequences. Such annotations can overlook the localized residue-level and geometric patterns encoded by sparse features. We introduce an automated and scalable method for interpreting SAE features in ESM-2 by using geometrically inspired features of the protein $\text{C}_α$ backbone. Across ESM-2 8M layers, an FDR-controlled discovery analysis shows that local geometry is significantly associated with many SAE features, with varying levels of predictive strength, expanding coverage beyond database and sequence-based methods. In particular, geometry can distinguish SAE features sharing the same database annotation, revealing substructure within known biological labels. A significant portion of SAE features activate on unannotated metagenomic protein sequences enabling us to use our SAE annotations to better understand these sequences. In addition, ablation experiments at the level of contact prediction show that removing found geometric features shifts ESM-2's predicted contact maps in the direction of the descriptor. This provides a robust method of annotating proteins activated within SAE neurons at a residue level, providing a bridge between mechanistic interpretability and structural biology. --- ### Redwood: A Frontier AI Accelerator Designed, Verified, and Deployed from Scratch in 2 Weeks by AI Authors: Architect Labs Published: 2026-08-26 Categories: cs.AR, cs.AI arXiv: https://arxiv.org/abs/2608.26418v2 Modern AI workloads and the hardware that runs them evolve on different timescales: architectural definition precedes volume silicon by years, while target workloads shift in months. Design decisions are therefore committed under deep uncertainty and paid for twice, once in the generality added as a hedge, and again when new workloads map poorly onto frozen silicon. As Moore's Law stagnates, specialization is the main remaining source of performance-per-watt and demands a design cycle that runs at the cadence of the workloads. We present an end-to-end AI system that collapses the software-to-silicon stack into a single optimization loop, where hardware and software are co-designed and verified under one objective. Its first demonstration is Redwood, a frontier AI accelerator built for single-batch, low-power, ultra-low-latency inference for physical AI. From a high-level specification by two human architects, the system autonomously generated the performance model, RTL design, UVM environments, formal proofs, firmware, and kernels in under two weeks with no human intervention below the specification. Every block reached 95% coverage via commercial EDA tools, our proprietary formal engine, and hardware-in-the-loop validation. Specification changes were reverified and redeployed to hardware in under 48 hours. Redwood Nano, its ultra-low-power FPGA variant, runs multi-billion-parameter models like Llama and Qwen. Projected onto Samsung 8 nm, the Jetson Orin Nano's process class, Redwood delivers 1.75x the throughput at 1.9x lower power, a 3.4x performance-per-watt gain against a measured Jetson baseline on the same models. Qwen running on Redwood also helped design next-generation Redwood, an early step toward recursive self-improvement. To our knowledge, this is the first production-worthy AI accelerator designed end-to-end by an AI system and running a modern AI model. --- ### Towards a universal meta-optics solver via large language models Authors: Huanshu Zhang, Lei Kang, Yuyan Chen, Luxiang Wang, Zhaolong Cao, Douglas H. Werner Published: 2026-08-26 Categories: physics.optics, cs.LG arXiv: https://arxiv.org/abs/2608.26417v1 Metasurface design increasingly requires fast models that can operate across structurally distinct device families, rather than retraining a separate surrogate for every geometry class. Conventional neural network surrogates often depend on fixed-dimensional descriptors, family-specific output formats, and repeated architecture tuning, which limits their scalability across heterogeneous meta-atoms. Here, we present a unified large language model (LLM) workflow for multi-family metasurface modeling and inverse-design. Geometries, design parameters, and optical response channels were converted into a shared instruction-following text format and used to fine-tune Gemma-2-9B across 8 metasurface families. Compared with single-family baselines, the joint model simultaneously predicted the optical responses of all metasurface families while reducing the MSE for each family by an average of 56.5%. The same representation was also used for inverse design. These results show that a shared sequence-based LLM interface can provide a practical route to cross-family metasurface design while reducing the need for task-specific surrogate architectures. --- ### Case2Flow: Bridging Patient Cases and Guideline Flowcharts through Multimodal Retrieval Authors: Jiale Wei, Yufan Chen, Alexander Jaus, Zdravko Marinov, Julian Friedrich, Simon Reiß, Jens Kleesiek, Rainer Stiefelhagen Published: 2026-08-26 Categories: cs.CL, cs.IR arXiv: https://arxiv.org/abs/2608.26414v1 Medical guidelines encode rich, evidence-based decision logic, yet the specific decision artifact a clinician needs is hard to locate within a guideline, let alone across guidelines covering plausible diseases and treatments. While guideline passages have supported end-to-end question answering, flowcharts remain largely underused in decision support despite their ability to encode actionable clinical pathways. We therefore introduce Case2Flow, a task designed to retrieve the most relevant guideline flowchart for a given patient case from a collection of guideline documents. To support it, we construct FlowAtlas, a curated corpus of 202 flowcharts extracted from 2,080 medical guidelines, together with a pipeline that synthesises 1,911 aligned case-flowchart pairs. Our evaluation of multimodal retrieval methods reveals systematic failure modes, including overreliance on keywords and spurious token-patch matches induced by uninformative background regions in flowcharts. Motivated by this, we propose CRISP, a training-free scoring method that sharpens late-interaction retrieval by suppressing uninformative patches, discounting ambiguous token matches, and incorporating bidirectional query-image alignment. CRISP improves Recall@1 by up to 18.71 percentage points, while a blinded physician assessment on published case narratives provides preliminary feasibility evidence beyond synthetic queries. --- ### Simultaneous Envy and Equitability Guarantees Authors: Hadi Hosseini, Shraddha Pathak, Lirong Xia, Chengkai Zhang Published: 2026-08-26 Categories: cs.GT, cs.AI, econ.TH arXiv: https://arxiv.org/abs/2608.26410v2 Recent work in fair division has focused on either simultaneously satisfying closely related fairness notions or achieving a single notion across the ex-ante and ex-post worlds. We study the compatibility of two fundamentally different fairness notions: envy-freeness and equitability. For indivisible goods-only and chores-only settings, we study the existence and complexity of simultaneously satisfying their relaxations, revealing sharp contrasts between the two settings. We show that EF1+EQ1 may fail to exist even for normalized binary goods: we construct an instance with 113 agents and 341 goods in which every agent approves exactly 165 goods, but no complete allocation satisfies both notions. Our main algorithmic result computes an EF1+EQ1 allocation for every normalized binary goods instance with at most seven agents. Thus, the smallest number of agents admitting a counterexample lies between 8 and 113, leaving the cases from 8 through 112 unresolved. In sharp contrast, binary chores admit the stronger EFX+EQX guarantee for any number of agents, even without normalization. We further initiate the study of cross-notion ex-ante and ex-post guarantees, asking whether randomized allocations can provide ex-ante guarantees for one notion while preserving ex-post guarantees for another. --- ### LowRankArena: A Standardized Evaluation Platform for SVD-Based LLM Compression Authors: Zishan Shao, Lixun Zhang, Kangning Cui, Wenhao Wu, Jinhee Kim, Yixiao Wang, Ting Jiang, Hancheng Ye, Qinsi Wang, Fan Yang, Danyang Zhuo, Yiran Chen et al. Published: 2026-08-26 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2608.26389v1 SVD-based low-rank compression has become a fast-growing direction for reducing the memory and computational cost of large language models (LLMs). However, meaningful comparison across existing studies remains difficult as prior evaluations use varied benchmarks, inconsistent ratios, and diverse setups, often failing to isolate low-rank effects from auxiliary techniques. As a result, it remains unclear whether reported gains reflect method-level improvements or differences in evaluation protocol. This lack of comparability highlights the need for a unified, reproducible evaluation platform. To address this problem, we present LowRankArena, a standardized evaluation platform for SVD-based LLM compression. LowRankArena unifies task versions, uniform-precision compression budgets, comparison regimes, and inference measurements, and provides a reproducible pipeline with over 3 TiB released compressed checkpoints. Using LowRankArena, our aligned audit of five representative SVD methods reveals that prior findings are highly conditional under standardized protocols: clear leaders and performance tiers shift across backbones and keep ratios, multiple-choice accuracy can hide large perplexity degradation, and nominal low-rank savings yield workload-dependent and often limited end-to-end speedups. Our code is available at: https://github.com/Zishan-Shao/lowrankarena.git. --- ### Co-Evolving Structured Knowledge and Reasoning in Language Models Authors: Ryan Thomas Noonan, Linxi Zhao, Menghan Xu, Akanksha Sarkar, Mihir Mishra, Dongyoung Go, Kilian Q. Weinberger, Yoav Artzi, Jennifer J. Sun Published: 2026-08-26 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2608.26386v2 Retrieval-augmented methods improve factual accuracy by grounding language models in external knowledge, but retrieving over unstructured text often introduces irrelevant context and offers limited control over the retrieved information. Structured knowledge bases offer a more controllable alternative, yet they are expensive to construct and often brittle to reason over. To address these limitations, we propose KBevo: a co-evolving framework that jointly learns to construct a structured knowledge base and reason over it for knowledge-intensive question answering. By optimizing both components end-to-end with QA outcome rewards, our method enables reasoning success to directly improve the quality of the constructed knowledge base. This leads to larger, better-connected knowledge structures with higher answer reachability, while also improving compositional factual reasoning and controllability compared to standard retrieval baselines. --- ### Why RAGs Hallucinate: Penalty-Aware Evaluation of Retrieval-Augmented Generation Systems with Knowledge-Gap Canaries Authors: Alden Do Rosario, Hussein Younes, Felipe Pires Published: 2026-08-26 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2608.26385v1 Volume-based accuracy rewards retrieval-augmented generation (RAG) systems for guessing: a system that answers everything outscores one that declines when its knowledge base cannot support an answer. Building on the confidence-target analysis of Kalai et al. (2025), we present a penalty-aware evaluation framework for deployed RAG products, combining (i) asymmetric scoring (correct +1, wrong -4, abstain 0), (ii) knowledge-gap canaries, questions whose answers are verifiably absent from the knowledge base, so that any answer constitutes ungrounded generation from parametric memory, and (iii) a failure-attribution pipeline that separates retrieval, generation, and abstention-policy failures. Applying the framework to three commercial RAG systems and a no-retrieval baseline on SimpleQA-Verified (1,000 questions x 3 repeats, graded blind by a cross-family three-judge panel with 98.9% unanimity), we find that accuracy when answering is closely clustered across systems (97.0-98.0%), while canary violation rates differ roughly sixfold (16.7% vs. 98.1%). The systems are separated less by what they answer correctly than by whether they answer at all when they should not, and penalty-aware scoring reorders the volume-based ranking accordingly; the reordering is stable across penalty settings from k=1 to k=9. All code, configurations, transcripts, and judge votes are released for independent audit. --- ### VIPER: An Expert-Curated Benchmark for Vision-Language Models in Veterinary Pathology Authors: Luca L. Weishaupt, Simone de Brot, Javier Asin, Llorenç Grau-Roma, Nic G. Reitsam, Andrew H. Song, Dongmin Bang, Stefan T. Kaluziak, Long Phi Le, Jakob Nikolas Kather, Faisal Mahmood, Guillaume Jaume Published: 2026-08-26 Categories: cs.CV arXiv: https://arxiv.org/abs/2608.26382v1 Pathology vision-language models are advancing rapidly, yet existing benchmarks remain focused on human tissue, particularly oncology, leaving non-human pathology largely unaddressed. This gap is especially important in toxicologic pathology, where microscopic tissue examination of laboratory animals is a core component of preclinical drug safety assessment. To address it, we introduce VIPER, the first expert-curated benchmark for vision-language model evaluation in toxicologic pathology. VIPER contains 1,251 questions associated with 419 H&E-stained rat histology images across seven organ systems, covering multiple-choice, KPrim, and free-text formats. All questions were curated and validated by board-certified veterinary pathologists. In total, we benchmarked 16 models, including two newly introduced veterinary-pathology models, seven human pathology-specialized models, and seven general-purpose frontier models. The results identify a substantial domain gap between veterinary and human pathology, expose the risk of over-diagnosis of normal tissue in frontier models, and show that domain-specific training remains critical for visually grounded predictions. VIPER data and evaluation code are available at https://github.com/mahmoodlab/viper. --- ### Assessing the Downstream Utility of Evidence-Aware Retrieval in RAG Authors: Utshab Kumar Ghosh, Debayan Mukhopadhyay, Shubham Chatterjee Published: 2026-08-26 Categories: cs.IR, cs.CL arXiv: https://arxiv.org/abs/2608.26379v1 Retrieval evaluation for retrieval-augmented generation (RAG) is increasingly designed around whether retrieved passages contain evidence that can support generation, rather than topical relevance alone. We study whether this closer alignment with downstream evidence needs also makes retrieval evaluation more useful for the decisions built from it. Across five retrieval benchmarks and an end-to-end TREC RAG 2025 setting, we examine an answer-support signal in four roles: comparing retrievers, guiding retrieval training and system selection, predicting downstream answer quality, and filtering the evidence supplied to a generator. The signal changes retrieval rankings, but its downstream value is not uniform. It does not reliably improve retriever training; the benefit of using it for system selection depends on how the generator is instructed to use the retrieved evidence; and retrieval scores based on it do not robustly predict answer quality on unseen topics. In a direct evidence intervention, human annotators confirm that filtering preferentially preserves passages containing useful answer evidence, yet different answer evaluators reach different conclusions about whether the resulting answers improve. These results show that making retrieval evaluation more closely reflect the evidence needed for generation does not by itself make every downstream use of that evaluation more reliable. RAG evaluation methods should therefore be assessed with respect to the particular comparisons, decisions, and conclusions they are intended to support. --- ### CG4AI: A Column Generation Framework for Training AI Models Under Constraints Authors: Youcef Magnouche, Abderrahmane Driouch, Sébastien Martin, Pierre Bauguion Published: 2026-08-26 Categories: cs.LG, cs.AI, cs.DM arXiv: https://arxiv.org/abs/2608.26375v1 Standard machine-learning training minimizes a loss function over a dataset, but does not guarantee that the resulting model will satisfy predefined rules or constraints on its outputs. In many real-world applications, ranging from autonomous systems to network routing, such guarantees are essential. We propose CG4AI, a framework that builds a convex combination of AI models while enforcing linear constraints on the combined output. A master linear program (LP) determines the optimal mixture weights, while a pricing subproblem generates new models guided by LP dual variables, focusing attention on the most violated constraints. A cutting-plane procedure extends feasibility guarantees beyond the training set. We apply CG4AI to two problems: (i) digit classification on MNIST, where we demonstrate four distinct uses of constraints, learning from constraints alone, improving adversarial robustness, correcting misclassified examples, and enforcing output relabeling; and (ii) the multi-commodity flow problem, where link capacity constraints are enforced on neural-network routing predictors. Experiments on MNIST and standard SNDLIB benchmark networks show that CG4AI reliably produces feasible predictors while achieving better accuracy than single-model baselines. ---