# Research Papers — 2026-01 · bundle page 102 of 189

> Full-text markdown of papers #5051–5100 of 9403 published in 2026-01, newest first. AI-readable mirrors of the full 9403-paper 2026-01 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-01/page/102
[← Previous bundle](https://aipod.dev/papers/archive/2026-01/page/101.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-01/page/103.md)

### Neural Chain-of-Thought Search: Searching the Optimal Reasoning Path to Enhance Large Language Models

Authors: Guoming Ling, Zhongzhan Huang, Yupei Lin, Junxin Li, Shanshan Zhong, Hefeng Wu, Liang Lin
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11340v2

Chain-of-Thought reasoning has significantly enhanced the problem-solving capabilities of Large Language Models. Unfortunately, current models generate reasoning steps sequentially without foresight, often becoming trapped in suboptimal reasoning paths with redundant steps. In contrast, we introduce Neural Chain-of-Thought Search (NCoTS), a framework that reformulates reasoning as a dynamic search for the optimal thinking strategy. By quantitatively characterizing the solution space, we reveal the existence of sparse superior reasoning paths that are simultaneously more accurate and concise than standard outputs. Our method actively navigates towards these paths by evaluating candidate reasoning operators using a dual-factor heuristic that optimizes for both correctness and computational cost. Consequently, NCoTS achieves a Pareto improvement across diverse reasoning benchmarks, boosting accuracy by over 3.5% while reducing generation length by over 22%. Our code and data are available at https://github.com/MilkThink-Lab/Neural-CoT-Search.

---

### Beer-Lambert Autoencoder for Unsupervised Stain Representation Learning and Deconvolution in Multi-immunohistochemical Brightfield Histology Images

Authors: Mark Eastwood, Thomas McKee, Zedong Hu, Sabine Tejpar, Fayyaz Minhas
Published: 2026-01-16
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2601.11336v1

Separating the contributions of individual chromogenic stains in RGB histology whole slide images (WSIs) is essential for stain normalization, quantitative assessment of marker expression, and cell-level readouts in immunohistochemistry (IHC). Classical Beer-Lambert (BL) color deconvolution is well-established for two- or three-stain settings, but becomes under-determined and unstable for multiplex IHC (mIHC) with K>3 chromogens. We present a simple, data-driven encoder-decoder architecture that learns cohort-specific stain characteristics for mIHC RGB WSIs and yields crisp, well-separated per-stain concentration maps. The encoder is a compact U-Net that predicts K nonnegative concentration channels; the decoder is a differentiable BL forward model with a learnable stain matrix initialized from typical chromogen hues. Training is unsupervised with a perceptual reconstruction objective augmented by loss terms that discourage unnecessary stain mixing. On a colorectal mIHC panel comprising 5 stains (H, CDX2, MUC2, MUC5, CD8) we show excellent RGB reconstruction, and significantly reduced inter-channel bleed-through compared with matrix-based deconvolution. Code and model are available at https://github.com/measty/StainQuant.git.

---

### Information Theoretic Perspective on Representation Learning

Authors: Deborah Pereg, Michael Wand
Published: 2026-01-16
Categories: cs.IT, cs.LG
arXiv: https://arxiv.org/abs/2601.11334v2

An information-theoretic framework is introduced to analyze last-layer embedding, focusing on learned representations for regression tasks. We define representation-rate and derive limits on the reliability with which input-output information can be represented as is inherently determined by the input-source entropy. We further define representation capacity in a perturbed setting, and representation rate-distortion for a compressed output. We derive the achievable capacity, the achievable representation-rate, and their converse. Finally, we combine the results in a unified setting.

---

### Idea First, Code Later: Disentangling Problem Solving from Code Generation in Evaluating LLMs for Competitive Programming

Authors: Sama Hadhoud, Alaa Elsetohy, Frederikus Hudi, Jan Christian Blaise Cruz, Steven Halim, Alham Fikri Aji
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11332v1

Large Language Models (LLMs) increasingly succeed on competitive programming problems, yet existing evaluations conflate algorithmic reasoning with code-level implementation. We argue that competitive programming is fundamentally a problem-solving task and propose centering natural-language editorials in both solution generation and evaluation. Generating an editorial prior to code improves solve rates for some LLMs, with substantially larger gains when using expertly written gold editorials. However, even with gold editorials, models continue to struggle with implementation, while the gap between generated and gold editorials reveals a persistent problem-solving bottleneck in specifying correct and complete algorithms. Beyond pass/fail metrics, we diagnose reasoning errors by comparing model-generated editorials to gold standards using expert annotations and validate an LLM-as-a-judge protocol for scalable evaluation. We introduce a dataset of 83 ICPC-style problems with gold editorials and full test suites, and evaluate 19 LLMs, arguing that future benchmarks should explicitly separate problem solving from implementation.

---

### F-Actor: Controllable Conversational Behaviour in Full-Duplex Models

Authors: Maike Züfle, Ondrej Klejch, Nicholas Sanders, Jan Niehues, Alexandra Birch, Tsz Kin Lam
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11329v3

Spoken conversational systems require more than accurate speech generation to have human-like conversations: to feel natural and engaging, they must produce conversational behaviour that adapts dynamically to the context. Current spoken conversational systems, however, rarely allow such customization, limiting their naturalness and usability. In this work, we present the first open, instruction-following full-duplex conversational speech model that can be trained efficiently under typical academic resource constraints. By keeping the audio encoder frozen and finetuning only the language model, our model requires just 2,000 hours of data, without relying on large-scale pretraining or multi-stage optimization. The model can follow explicit instructions to control speaker voice, conversation topic, conversational behaviour (e.g., backchanneling and interruptions), and dialogue initiation. We propose a single-stage training protocol and systematically analyze design choices. Both the model and training code is released to enable reproducible research on controllable full-duplex speech systems.

---

### GENPACK: KPI-Guided Multi-Criteria Genetic Algorithm for Industrial 3D Bin Packing

Authors: Dheeraj Poolavaram, Carsten Markgraf, Sebastian Dorn
Published: 2026-01-16
Categories: cs.NE
arXiv: https://arxiv.org/abs/2601.11325v3

The three-dimensional bin packing problem (3D-BPP) is a longstanding challenge in operations research and logistics. While classical heuristics and constructive methods can generate packings efficiently, they often fail to satisfy industrial requirements such as stability, balance, and handling feasibility. Metaheuristics such as genetic algorithms (GAs) offer greater flexibility, but pure GA approaches frequently struggle with efficiency, parameter sensitivity, and scalability to industrial order sizes. These limitations are particularly evident at real-world pallet dimensions, where even state-of-the-art methods often fail to produce robust, deployable solutions. We propose a KPI-guided GA-based pipeline for industrial 3D-BPP that integrates key performance indicators (KPIs) directly into a scalarized fitness function. The method combines a layer-based chromosome representation, domain-specific operators, and constructive heuristics to balance efficiency and feasibility. On the BED-BPP benchmark of 1,500 real-world orders, our GENPACK pipeline consistently outperforms heuristic and learning-based baselines, achieving up to 35% higher space utilization and 15-20% stronger surface support, while exhibiting lower variance across orders. These gains come at a modest runtime cost but remain practical for batch-scale deployment, yielding stable, balanced, and space-efficient packings.

---

### Enhancing Vision Language Models with Logic Reasoning for Situational Awareness

Authors: Pavana Pradeep, Krishna Kant, Suya Yu
Published: 2026-01-16
Categories: cs.CV, cs.LO
arXiv: https://arxiv.org/abs/2601.11322v1

Vision-Language Models (VLMs) offer the ability to generate high-level, interpretable descriptions of complex activities from images and videos, making them valuable for situational awareness (SA) applications. In such settings, the focus is on identifying infrequent but significant events with high reliability and accuracy, while also extracting fine-grained details and assessing recognition quality. In this paper, we propose an approach that integrates VLMs with traditional computer vision methods through explicit logic reasoning to enhance SA in three key ways: (a) extracting fine-grained event details, (b) employing an intelligent fine-tuning (FT) strategy that achieves substantially higher accuracy than uninformed selection, and (c) generating justifications for VLM outputs during inference. We demonstrate that our intelligent FT mechanism improves the accuracy and provides a valuable means, during inferencing, to either confirm the validity of the VLM output or indicate why it may be questionable.

---

### Membership Inference on LLMs in the Wild

Authors: Jiatong Yi, Yanyang Li
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11314v1

Membership Inference Attacks (MIAs) act as a crucial auditing tool for the opaque training data of Large Language Models (LLMs). However, existing techniques predominantly rely on inaccessible model internals (e.g., logits) or suffer from poor generalization across domains in strict black-box settings where only generated text is available. In this work, we propose SimMIA, a robust MIA framework tailored for this text-only regime by leveraging an advanced sampling strategy and scoring mechanism. Furthermore, we present WikiMIA-25, a new benchmark curated to evaluate MIA performance on modern proprietary LLMs. Experiments demonstrate that SimMIA achieves state-of-the-art results in the black-box setting, rivaling baselines that exploit internal model information.

---

### FORESTLLM: Large Language Models Make Random Forest Great on Few-shot Tabular Learning

Authors: Zhihan Yang, Jiaqi Wei, Xiang Zhang, Haoyu Dong, Yiwen Wang, Xiaoke Guo, Pengkun Zhang, Yiwei Xu, Chenyu You
Published: 2026-01-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.11311v1

Tabular data high-stakes critical decision-making in domains such as finance, healthcare, and scientific discovery. Yet, learning effectively from tabular data in few-shot settings, where labeled examples are scarce, remains a fundamental challenge. Traditional tree-based methods often falter in these regimes due to their reliance on statistical purity metrics, which become unstable and prone to overfitting with limited supervision. At the same time, direct applications of large language models (LLMs) often overlook its inherent structure, leading to suboptimal performance. To overcome these limitations, we propose FORESTLLM, a novel framework that unifies the structural inductive biases of decision forests with the semantic reasoning capabilities of LLMs. Crucially, FORESTLLM leverages the LLM only during training, treating it as an offline model designer that encodes rich, contextual knowledge into a lightweight, interpretable forest model, eliminating the need for LLM inference at test time. Our method is two-fold. First, we introduce a semantic splitting criterion in which the LLM evaluates candidate partitions based on their coherence over both labeled and unlabeled data, enabling the induction of more robust and generalizable tree structures under few-shot supervision. Second, we propose a one-time in-context inference mechanism for leaf node stabilization, where the LLM distills the decision path and its supporting examples into a concise, deterministic prediction, replacing noisy empirical estimates with semantically informed outputs. Across a diverse suite of few-shot classification and regression benchmarks, FORESTLLM achieves state-of-the-art performance.

---

### Context-Aware Semantic Segmentation via Stage-Wise Attention

Authors: Antoine Carreaud, Elias Naha, Arthur Chansel, Nina Lahellec, Jan Skaloud, Adrien Gressin
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11310v2

Semantic ultra-high-resolution (UHR) image segmentation is essential in remote sensing applications such as aerial mapping and environmental monitoring. Transformer-based models remain challenging in this setting because memory grows quadratically with the number of tokens, limiting either spatial resolution or contextual scope. We introduce CASWiT (Context-Aware Stage-Wise Transformer), a dual-branch Swin-based architecture that injects low-resolution contextual information into fine-grained high-resolution features through lightweight stage-wise cross-attention. To strengthen cross-scale learning, we also propose a SimMIM-style pretraining strategy based on masked reconstruction of the high-resolution image. Extensive experiments on the large-scale FLAIR-HUB aerial dataset demonstrate the effectiveness of CASWiT. Under our RGB-only UHR protocol, CASWiT reaches 66.37% mIoU with a SegFormer decoder, improving over strong RGB baselines while also improving boundary quality. On the URUR benchmark, CASWiT reaches 49.2% mIoU under the official evaluation protocol, and it also transfers effectively to medical UHR segmentation benchmarks. Code and pretrained models are available at https://huggingface.co/collections/heig-vd-geo/caswit

---

### SAMannot: A Memory-Efficient, Local, Open-source Framework for Interactive Video Instance Segmentation based on SAM2

Authors: Gergely Dinya, András Gelencsér, Krisztina Kupán, Clemens Küpper, Kristóf Karacs, Anna Gelencsér-Horváth
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11301v2

Current research workflows for precise video segmentation are often forced into a compromise between labor-intensive manual curation, costly commercial platforms, and/or privacy-compromising cloud-based services. The demand for high-fidelity video instance segmentation in research is often hindered by the bottleneck of manual annotation and the privacy concerns of cloud-based tools. We present SAMannot, an open-source, local framework that integrates the Segment Anything Model 2 (SAM2) into a human-in-the-loop workflow. To address the high resource requirements of foundation models, we modified the SAM2 dependency and implemented a processing layer that minimizes computational overhead and maximizes throughput, ensuring a highly responsive user interface. Key features include persistent instance identity management, an automated ``lock-and-refine'' workflow with barrier frames, and a mask-skeletonization-based auto-prompting mechanism. SAMannot facilitates the generation of research-ready datasets in YOLO and PNG formats alongside structured interaction logs. Verified through animal behavior tracking use-cases and subsets of the LVOS and DAVIS benchmark datasets, the tool provides a scalable, private, and cost-effective alternative to commercial platforms for complex video annotation tasks.

---

### One LLM to Train Them All: Multi-Task Learning Framework for Fact-Checking

Authors: Malin Astrid Larsson, Harald Fosen Grunnaleite, Vinay Setty
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11293v1

Large language models (LLMs) are reshaping automated fact-checking (AFC) by enabling unified, end-to-end verification pipelines rather than isolated components. While large proprietary models achieve strong performance, their closed weights, complexity, and high costs limit sustainability. Fine-tuning smaller open weight models for individual AFC tasks can help but requires multiple specialized models resulting in high costs. We propose \textbf{multi-task learning (MTL)} as a more efficient alternative that fine-tunes a single model to perform claim detection, evidence ranking, and stance detection jointly. Using small decoder-only LLMs (e.g., Qwen3-4b), we explore three MTL strategies: classification heads, causal language modeling heads, and instruction-tuning, and evaluate them across model sizes, task orders, and standard non-LLM baselines. While multitask models do not universally surpass single-task baselines, they yield substantial improvements, achieving up to \textbf{44\%}, \textbf{54\%}, and \textbf{31\%} relative gains for claim detection, evidence re-ranking, and stance detection, respectively, over zero-/few-shot settings. Finally, we also provide practical, empirically grounded guidelines to help practitioners apply MTL with LLMs for automated fact-checking.

---

### Efficient On-Board Processing of Oblique UAV Video for Rapid Flood Extent Mapping

Authors: Vishisht Sharma, Sam Leroux, Lisa Landuyt, Nick Witvrouwen, Pieter Simoens
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11290v1

Effective disaster response relies on rapid disaster response, where oblique aerial video is the primary modality for initial scouting due to its ability to maximize spatial coverage and situational awareness in limited flight time. However, the on-board processing of high-resolution oblique streams is severely bottlenecked by the strict Size, Weight, and Power (SWaP) constraints of Unmanned Aerial Vehicles (UAVs). The computational density required to process these wide-field-of-view streams precludes low-latency inference on standard edge hardware. To address this, we propose Temporal Token Reuse (TTR), an adaptive inference framework capable of accelerating video segmentation on embedded devices. TTR exploits the intrinsic spatiotemporal redundancy of aerial video by formulating image patches as tokens; it utilizes a lightweight similarity metric to dynamically identify static regions and propagate their precomputed deep features, thereby bypassing redundant backbone computations. We validate the framework on standard benchmarks and a newly curated Oblique Floodwater Dataset designed for hydrological monitoring. Experimental results on edge-grade hardware demonstrate that TTR achieves a 30% reduction in inference latency with negligible degradation in segmentation accuracy (< 0.5% mIoU). These findings confirm that TTR effectively shifts the operational Pareto frontier, enabling high-fidelity, real-time oblique video understanding for time-critical remote sensing missions

---

### XChoice: Explainable Evaluation of AI-Human Alignment in LLM-based Constrained Choice Decision Making

Authors: Weihong Qi, Fan Huang, Rasika Muralidharan, Jisun An, Haewoon Kwak
Published: 2026-01-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.11286v1

We present XChoice, an explainable framework for evaluating AI-human alignment in constrained decision making. Moving beyond outcome agreement such as accuracy and F1 score, XChoice fits a mechanism-based decision model to human data and LLM-generated decisions, recovering interpretable parameters that capture the relative importance of decision factors, constraint sensitivity, and implied trade-offs. Alignment is assessed by comparing these parameter vectors across models, options, and subgroups. We demonstrate XChoice on Americans' daily time allocation using the American Time Use Survey (ATUS) as human ground truth, revealing heterogeneous alignment across models and activities and salient misalignment concentrated in Black and married groups. We further validate robustness of XChoice via an invariance analysis and evaluate targeted mitigation with a retrieval augmented generation (RAG) intervention. Overall, XChoice provides mechanism-based metrics that diagnose misalignment and support informed improvements beyond surface outcome matching.

---

### Metabolomic Biomarker Discovery for ADHD Diagnosis Using Interpretable Machine Learning

Authors: Nabil Belacel, Mohamed Rachid Boulassel
Published: 2026-01-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.11283v2

Attention Deficit Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder with limited objective diagnostic tools, highlighting the urgent need for objective, biology-based diagnostic frameworks in precision psychiatry. We integrate urinary metabolomics with an interpretable machine learning framework to identify biochemical signatures associated with ADHD. Targeted metabolomic profiles from 52 ADHD and 46 control participants were analyzed using a Closest Resemblance (CR) classifier with embedded feature selection. The CR model outperformed Random Forest and K-Nearest Neighbor classifiers, achieving an AUC > 0.97 based on a reduced panel of 14 metabolites. These metabolites including dopamine 4-sulfate, N-acetylaspartylglutamic acid, and citrulline map to dopaminergic neurotransmission and amino acid metabolism pathways, offering mechanistic insight into ADHD pathophysiology. The CR classifier's transparent decision boundaries and low computational cost support integration into targeted metabolomic assays and future point of care diagnostic platforms. Overall, this work demonstrates a translational framework combining metabolomics and interpretable machine learning to advance objective, biologically informed diagnostic strategies for ADHD.

---

### From SERPs to Sound: How Search Engine Result Pages and AI-generated Podcasts Interact to Influence User Attitudes on Controversial Topics

Authors: Junjie Wang, Gaole He, Alisa Rieger, Ujwal Gadiraju
Published: 2026-01-16
Categories: cs.IR, cs.AI, cs.HC, cs.SI
arXiv: https://arxiv.org/abs/2601.11282v1

Compared to search engine result pages (SERPs), AI-generated podcasts represent a relatively new and relatively more passive modality of information consumption, delivering narratives in a naturally engaging format. As these two media increasingly converge in everyday information-seeking behavior, it is essential to explore how their interaction influences user attitudes, particularly in contexts involving controversial, value-laden, and often debated topics. Addressing this need, we aim to understand how information mediums of present-day SERPs and AI-generated podcasts interact to shape the opinions of users. To this end, through a controlled user study (N=483), we investigated user attitudinal effects of consuming information via SERPs and AI-generated podcasts, focusing on how the sequence and modality of exposure shape user opinions. A majority of users in our study corresponded to attitude change outcomes, and we found an effect of sequence on attitude change. Our results further revealed a role of viewpoint bias and the degree of topic controversiality in shaping attitude change, although we found no effect of individual moderators.

---

### X-Distill: Cross-Architecture Vision Distillation for Visuomotor Learning

Authors: Maanping Shao, Feihong Zhang, Gu Zhang, Baiye Cheng, Zhengrong Xue, Huazhe Xu
Published: 2026-01-16
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2601.11269v1

Visuomotor policies often leverage large pre-trained Vision Transformers (ViTs) for their powerful generalization capabilities. However, their significant data requirements present a major challenge in the data-scarce context of most robotic learning settings, where compact CNNs with strong inductive biases can be more easily optimized. To address this trade-off, we introduce X-Distill, a simple yet highly effective method that synergizes the strengths of both architectures. Our approach involves an offline, cross-architecture knowledge distillation, transferring the rich visual representations of a large, frozen DINOv2 teacher to a compact ResNet-18 student on the general-purpose ImageNet dataset. This distilled encoder, now endowed with powerful visual priors, is then jointly fine-tuned with a diffusion policy head on the target manipulation tasks. Extensive experiments on $34$ simulated benchmarks and $5$ challenging real-world tasks demonstrate that our method consistently outperforms policies equipped with from-scratch ResNet or fine-tuned DINOv2 encoders. Notably, X-Distill also surpasses 3D encoders that utilize privileged point cloud observations or much larger Vision-Language Models. Our work highlights the efficacy of a simple, well-founded distillation strategy for achieving state-of-the-art performance in data-efficient robotic manipulation.

---

### Sample-Near-Optimal Agnostic Boosting with Improved Running Time

Authors: Arthur da Cunha, Mikael Møller Høgsgaard, Andrea Paudice
Published: 2026-01-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.11265v3

Boosting is a powerful method that turns weak learners, which perform only slightly better than random guessing, into strong learners with high accuracy. While boosting is well understood in the classic setting, it is less so in the agnostic case, where no assumptions are made about the data. Indeed, only recently was the sample complexity of agnostic boosting nearly settled arXiv:2503.09384, but the known algorithm achieving this bound has exponential running time. In this work, we propose the first agnostic boosting algorithm with near-optimal sample complexity, running in time polynomial in the sample size when considering the other parameters of the problem fixed.

---

### Scalable Music Cover Retrieval Using Lyrics-Aligned Audio Embeddings

Authors: Joanne Affolter, Benjamin Martin, Elena V. Epure, Gabriel Meseguer-Brocal, Frédéric Kaplan
Published: 2026-01-16
Categories: cs.SD, cs.IR, cs.LG
arXiv: https://arxiv.org/abs/2601.11262v2

Music Cover Retrieval, also known as Version Identification, aims to recognize distinct renditions of the same underlying musical work, a task central to catalog management, copyright enforcement, and music retrieval. State-of-the-art approaches have largely focused on harmonic and melodic features, employing increasingly complex audio pipelines designed to be invariant to musical attributes that often vary widely across covers. While effective, these methods demand substantial training time and computational resources. By contrast, lyrics constitute a strong invariant across covers, though their use has been limited by the difficulty of extracting them accurately and efficiently from polyphonic audio. Early methods relied on simple frameworks that limited downstream performance, while more recent systems deliver stronger results but require large models integrated within complex multimodal architectures. We introduce LIVI (Lyrics-Informed Version Identification), an approach that seeks to balance retrieval accuracy with computational efficiency. First, LIVI leverages supervision from state-of-the-art transcription and text embedding models during training to achieve retrieval accuracy on par with--or superior to--harmonic-based systems. Second, LIVI remains lightweight and efficient by removing the transcription step at inference, challenging the dominance of complexity-heavy pipelines.

---

### Effects of Introducing Synaptic Scaling on Spiking Neural Network Learning

Authors: Shinnosuke Touda, Hirotsugu Okuno
Published: 2026-01-16
Categories: cs.NE, cs.LG
arXiv: https://arxiv.org/abs/2601.11261v1

Spiking neural networks (SNNs) employing unsupervised learning methods inspired by neural plasticity are expected to be a new framework for artificial intelligence. In this study, we investigated the effect of multiple types of neural plasticity, such as spike-time-dependent plasticity (STDP) and synaptic scaling, on the learning in a winner-take-all (WTA) network composed of spiking neurons. We implemented a WTA network with multiple types of neural plasticity using Python. The MNIST and the Fashion-MNIST datasets were used for training and testing. We varied the number of neurons, the time constant of STDP, and the normalization method used in synaptic scaling to compare classification accuracy. The results demonstrated that synaptic scaling based on the L2 norm was the most effective in improving classification performance. By implementing L2-norm-based synaptic scaling and setting the number of neurons in both excitatory and inhibitory layers to 400, the network achieved classification accuracies of 88.84 % on the MNIST dataset and 68.01 % on the Fashion-MNIST dataset after one epoch of training.

---

### Latent Dynamics Graph Convolutional Networks for model order reduction of parameterized time-dependent PDEs

Authors: Lorenzo Tomada, Federico Pichi, Gianluigi Rozza
Published: 2026-01-16
Categories: cs.LG, math.NA
arXiv: https://arxiv.org/abs/2601.11259v1

Graph Neural Networks (GNNs) are emerging as powerful tools for nonlinear Model Order Reduction (MOR) of time-dependent parameterized Partial Differential Equations (PDEs). However, existing methodologies struggle to combine geometric inductive biases with interpretable latent behavior, overlooking dynamics-driven features or disregarding spatial information. In this work, we address this gap by introducing Latent Dynamics Graph Convolutional Network (LD-GCN), a purely data-driven, encoder-free architecture that learns a global, low-dimensional representation of dynamical systems conditioned on external inputs and parameters. The temporal evolution is modeled in the latent space and advanced through time-stepping, allowing for time-extrapolation, and the trajectories are consistently decoded onto geometrically parameterized domains using a GNN. Our framework enhances interpretability by enabling the analysis of the reduced dynamics and supporting zero-shot prediction through latent interpolation. The methodology is mathematically validated via a universal approximation theorem for encoder-free architectures, and numerically tested on complex computational mechanics problems involving physical and geometric parameters, including the detection of bifurcating phenomena for Navier-Stokes equations. Code availability: https://github.com/lorenzotomada/ld-gcn-rom

---

### Knowledge is Not Enough: Injecting RL Skills for Continual Adaptation

Authors: Pingzhi Tang, Yiding Wang, Muhan Zhang
Published: 2026-01-16
Categories: cs.LG, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2601.11258v2

Large Language Models (LLMs) face the "knowledge cutoff" challenge, where their frozen parametric memory prevents direct internalization of new information. While Supervised Fine-Tuning (SFT) is commonly used to update model knowledge, it often updates factual content without reliably improving the model's ability to use the newly incorporated information for question answering or decision-making. Reinforcement Learning (RL) is essential for acquiring reasoning skills; however, its high computational cost makes it impractical for efficient online adaptation. We empirically observe that the parameter updates induced by SFT and RL are nearly orthogonal. Based on this observation, we propose Parametric Skill Transfer (PaST), a framework that supports modular skill transfer for efficient and effective knowledge adaptation. By extracting a domain-agnostic Skill Vector from a source domain, we can linearly inject knowledge manipulation skills into a target model after it has undergone lightweight SFT on new data. Experiments on knowledge-incorporation QA (SQuAD, LooGLE) and agentic tool-use benchmarks (ToolBench) demonstrate the effectiveness of our method. On SQuAD, PaST outperforms the state-of-the-art self-editing SFT baseline by up to 9.9 points. PaST further scales to long-context QA on LooGLE with an 8.0-point absolute accuracy gain, and improves zero-shot ToolBench success rates by +10.3 points on average with consistent gains across tool categories, indicating strong scalability and cross-domain transferability of the Skill Vector.

---

### Reasoning in Trees: Improving Retrieval-Augmented Generation for Multi-Hop Question Answering

Authors: Yuling Shi, Maolin Sun, Zijun Liu, Mo Yang, Yixiong Fang, Tianran Sun, Xiaodong Gu
Published: 2026-01-16
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2601.11255v1

Retrieval-Augmented Generation (RAG) has demonstrated significant effectiveness in enhancing large language models (LLMs) for complex multi-hop question answering (QA). For multi-hop QA tasks, current iterative approaches predominantly rely on LLMs to self-guide and plan multi-step exploration paths during retrieval, leading to substantial challenges in maintaining reasoning coherence across steps from inaccurate query decomposition and error propagation. To address these issues, we introduce Reasoning Tree Guided RAG (RT-RAG), a novel hierarchical framework for complex multi-hop QA. RT-RAG systematically decomposes multi-hop questions into explicit reasoning trees, minimizing inaccurate decomposition through structured entity analysis and consensus-based tree selection that clearly separates core queries, known entities, and unknown entities. Subsequently, a bottom-up traversal strategy employs iterative query rewriting and refinement to collect high-quality evidence, thereby mitigating error propagation. Comprehensive experiments show that RT-RAG substantially outperforms state-of-the-art methods by 7.0% F1 and 6.0% EM, demonstrating the effectiveness of RT-RAG in complex multi-hop QA.

---

### FTDMamba: Frequency-Assisted Temporal Dilation Mamba for Unmanned Aerial Vehicle Video Anomaly Detection

Authors: Cheng-Zhuang Liu, Si-Bao Chen, Qing-Ling Shu, Chris Ding, Jin Tang, Bin Luo
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11254v1

Recent advances in video anomaly detection (VAD) mainly focus on ground-based surveillance or unmanned aerial vehicle (UAV) videos with static backgrounds, whereas research on UAV videos with dynamic backgrounds remains limited. Unlike static scenarios, dynamically captured UAV videos exhibit multi-source motion coupling, where the motion of objects and UAV-induced global motion are intricately intertwined. Consequently, existing methods may misclassify normal UAV movements as anomalies or fail to capture true anomalies concealed within dynamic backgrounds. Moreover, many approaches do not adequately address the joint modeling of inter-frame continuity and local spatial correlations across diverse temporal scales. To overcome these limitations, we propose the Frequency-Assisted Temporal Dilation Mamba (FTDMamba) network for UAV VAD, including two core components: (1) a Frequency Decoupled Spatiotemporal Correlation Module, which disentangles coupled motion patterns and models global spatiotemporal dependencies through frequency analysis; and (2) a Temporal Dilation Mamba Module, which leverages Mamba's sequence modeling capability to jointly learn fine-grained temporal dynamics and local spatial structures across multiple temporal receptive fields. Additionally, unlike existing UAV VAD datasets which focus on static backgrounds, we construct a large-scale Moving UAV VAD dataset (MUVAD), comprising 222,736 frames with 240 anomaly events across 12 anomaly types. Extensive experiments demonstrate that FTDMamba achieves state-of-the-art (SOTA) performance on two public static benchmarks and the new MUVAD dataset. The code and MUVAD dataset will be available at: https://github.com/uavano/FTDMamba.

---

### Beyond Model Scaling: Test-Time Intervention for Efficient Deep Reasoning

Authors: Qianyue Wang, Jinwu Hu, Yufeng Wang, Huanxiang Lin, Bolin Chen, Zhiquan Wen, Yaofo Chen, Mingkui Tan
Published: 2026-01-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.11252v1

Large Reasoning Models (LRMs) excel at multi-step reasoning but often suffer from inefficient reasoning processes like overthinking and overshoot, where excessive or misdirected reasoning increases computational cost and degrades performance. Existing efficient reasoning methods operate in a closed-loop manner, lacking mechanisms for external intervention to guide the reasoning process. To address this, we propose Think-with-Me, a novel test-time interactive reasoning paradigm that introduces external feedback intervention into the reasoning process. Our key insights are that transitional conjunctions serve as natural points for intervention, signaling phases of self-validation or exploration and using transitional words appropriately to prolong the reasoning enhances performance, while excessive use affects performance. Building on these insights, Think-with-Me pauses reasoning at these points for external feedback, adaptively extending or terminating reasoning to reduce redundancy while preserving accuracy. The feedback is generated via a multi-criteria evaluation (rationality and completeness) and comes from either human or LLM proxies. We train the target model using Group Relative Policy Optimization (GRPO) to adapt to this interactive mode. Experiments show that Think-with-Me achieves a superior balance between accuracy and reasoning length under limited context windows. On AIME24, Think-with-Me outperforms QwQ-32B by 7.19% in accuracy while reducing average reasoning length by 81% under an 8K window. The paradigm also benefits security and creative tasks.

---

### Language-Agnostic Visual Embeddings for Cross-Script Handwriting Retrieval

Authors: Fangke Chen, Tianhao Dong, Sirry Chen, Guobin Zhang, Yishu Zhang, Yining Chen
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11248v1

Handwritten word retrieval is vital for digital archives but remains challenging due to large handwriting variability and cross-lingual semantic gaps. While large vision-language models offer potential solutions, their prohibitive computational costs hinder practical edge deployment. To address this, we propose a lightweight asymmetric dual-encoder framework that learns unified, style-invariant visual embeddings. By jointly optimizing instance-level alignment and class-level semantic consistency, our approach anchors visual embeddings to language-agnostic semantic prototypes, enforcing invariance across scripts and writing styles. Experiments show that our method outperforms 28 baselines and achieves state-of-the-art accuracy on within-language retrieval benchmarks. We further conduct explicit cross-lingual retrieval, where the query language differs from the target language, to validate the effectiveness of the learned cross-lingual representations. Achieving strong performance with only a fraction of the parameters required by existing models, our framework enables accurate and resource-efficient cross-script handwriting retrieval.

---

### Image-Text Knowledge Modeling for Unsupervised Multi-Scenario Person Re-Identification

Authors: Zhiqi Pang, Lingling Zhao, Yang Liu, Chunyu Wang, Gaurav Sharma
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11243v1

We propose unsupervised multi-scenario (UMS) person re-identification (ReID) as a new task that expands ReID across diverse scenarios (cross-resolution, clothing change, etc.) within a single coherent framework. To tackle UMS-ReID, we introduce image-text knowledge modeling (ITKM) -- a three-stage framework that effectively exploits the representational power of vision-language models. We start with a pre-trained CLIP model with an image encoder and a text encoder. In Stage I, we introduce a scenario embedding in the image encoder and fine-tune the encoder to adaptively leverage knowledge from multiple scenarios. In Stage II, we optimize a set of learned text embeddings to associate with pseudo-labels from Stage I and introduce a multi-scenario separation loss to increase the divergence between inter-scenario text representations. In Stage III, we first introduce cluster-level and instance-level heterogeneous matching modules to obtain reliable heterogeneous positive pairs (e.g., a visible image and an infrared image of the same person) within each scenario. Next, we propose a dynamic text representation update strategy to maintain consistency between text and image supervision signals. Experimental results across multiple scenarios demonstrate the superiority and generalizability of ITKM; it not only outperforms existing scenario-specific methods but also enhances overall performance by integrating knowledge from multiple scenarios.

---

### Bio-inspired fine-tuning for selective transfer learning in image classification

Authors: Ana Davila, Jacinto Colan, Yasuhisa Hasegawa
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11235v1

Deep learning has significantly advanced image analysis across diverse domains but often depends on large, annotated datasets for success. Transfer learning addresses this challenge by utilizing pre-trained models to tackle new tasks with limited labeled data. However, discrepancies between source and target domains can hinder effective transfer learning. We introduce BioTune, a novel adaptive fine-tuning technique utilizing evolutionary optimization. BioTune enhances transfer learning by optimally choosing which layers to freeze and adjusting learning rates for unfrozen layers. Through extensive evaluation on nine image classification datasets, spanning natural and specialized domains such as medical imaging, BioTune demonstrates superior accuracy and efficiency over state-of-the-art fine-tuning methods, including AutoRGN and LoRA, highlighting its adaptability to various data characteristics and distribution changes. Additionally, BioTune consistently achieves top performance across four different CNN architectures, underscoring its flexibility. Ablation studies provide valuable insights into the impact of BioTune's key components on overall performance. The source code is available at https://github.com/davilac/BioTune.

---

### How DDAIR you? Disambiguated Data Augmentation for Intent Recognition

Authors: Galo Castillo-López, Alexis Lombard, Nasredine Semmar, Gaël de Chalendar
Published: 2026-01-16
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2601.11234v1

Large Language Models (LLMs) are effective for data augmentation in classification tasks like intent detection. In some cases, they inadvertently produce examples that are ambiguous with regard to untargeted classes. We present DDAIR (Disambiguated Data Augmentation for Intent Recognition) to mitigate this problem. We use Sentence Transformers to detect ambiguous class-guided augmented examples generated by LLMs for intent recognition in low-resource scenarios. We identify synthetic examples that are semantically more similar to another intent than to their target one. We also provide an iterative re-generation method to mitigate such ambiguities. Our findings show that sentence embeddings effectively help to (re)generate less ambiguous examples, and suggest promising potential to improve classification performance in scenarios where intents are loosely or broadly defined.

---

### FactCorrector: A Graph-Inspired Approach to Long-Form Factuality Correction of Large Language Models

Authors: Javier Carnerero-Cano, Massimiliano Pronesti, Radu Marinescu, Tigran Tchrakian, James Barry, Jasmina Gajcin, Yufang Hou, Alessandra Pascale, Elizabeth Daly
Published: 2026-01-16
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.11232v1

Large language models (LLMs) are widely used in knowledge-intensive applications but often generate factually incorrect responses. A promising approach to rectify these flaws is correcting LLMs using feedback. Therefore, in this paper, we introduce FactCorrector, a new post-hoc correction method that adapts across domains without retraining and leverages structured feedback about the factuality of the original response to generate a correction. To support rigorous evaluations of factuality correction methods, we also develop the VELI5 benchmark, a novel dataset containing systematically injected factual errors and ground-truth corrections. Experiments on VELI5 and several popular long-form factuality datasets show that the FactCorrector approach significantly improves factual precision while preserving relevance, outperforming strong baselines. We release our code at https://ibm.biz/factcorrector.

---

### Language of Thought Shapes Output Diversity in Large Language Models

Authors: Shaoyang Xu, Wenxuan Zhang
Published: 2026-01-16
Categories: cs.CL, cs.CY
arXiv: https://arxiv.org/abs/2601.11227v2

Output diversity is crucial for Large Language Models as it underpins pluralism and creativity. In this work, we reveal that controlling the language used during model thinking-the language of thought-provides a novel and structural source of output diversity. Our preliminary study shows that different thinking languages occupy distinct regions in a model's thinking space. Based on this observation, we study two repeated sampling strategies under multilingual thinking-Single-Language Sampling and Mixed-Language Sampling-and conduct diversity evaluation on outputs that are controlled to be in English, regardless of the thinking language used. Across extensive experiments, we demonstrate that switching the thinking language from English to non-English languages consistently increases output diversity, with a clear and consistent positive correlation such that languages farther from English in the thinking space yield larger gains. We further show that aggregating samples across multiple thinking languages yields additional improvements through compositional effects, and that scaling sampling with linguistic heterogeneity expands the model's diversity ceiling. Finally, we show that these findings translate into practical benefits in pluralistic alignment scenarios, leading to broader coverage of cultural knowledge and value orientations in LLM outputs. Our code is publicly available at https://github.com/iNLP-Lab/Multilingual-LoT-Diversity.

---

### Operator learning on domain boundary through combining fundamental solution-based artificial data and boundary integral techniques

Authors: Haochen Wu, Heng Wu, Benzhuo Lu
Published: 2026-01-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.11222v1

For linear partial differential equations with known fundamental solutions, this work introduces a novel operator learning framework that relies exclusively on domain boundary data, including solution values and normal derivatives, rather than full-domain sampling. By integrating the previously developed Mathematical Artificial Data (MAD) method, which enforces physical consistency, all training data are synthesized directly from the fundamental solutions of the target problems, resulting in a fully data-driven pipeline without the need for external measurements or numerical simulations. We refer to this approach as the Mathematical Artificial Data Boundary Neural Operator (MAD-BNO), which learns boundary-to-boundary mappings using MAD-generated Dirichlet-Neumann data pairs. Once trained, the interior solution at arbitrary locations can be efficiently recovered through boundary integral formulations, supporting Dirichlet, Neumann, and mixed boundary conditions as well as general source terms. The proposed method is validated on benchmark operator learning tasks for two-dimensional Laplace, Poisson, and Helmholtz equations, where it achieves accuracy comparable to or better than existing neural operator approaches while significantly reducing training time. The framework is naturally extensible to three-dimensional problems and complex geometries.

---

### MultiCaption: Detecting disinformation using multilingual visual claims

Authors: Rafael Martins Frade, Rrubaa Panchendrarajan, Arkaitz Zubiaga
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11220v1

Online disinformation poses an escalating threat to society, driven increasingly by the rapid spread of misleading content across both multimedia and multilingual platforms. While automated fact-checking methods have advanced in recent years, their effectiveness remains constrained by the scarcity of datasets that reflect these real-world complexities. To address this gap, we first present MultiCaption, a new dataset specifically designed for detecting contradictions in visual claims. Pairs of claims referring to the same image or video were labeled through multiple strategies to determine whether they contradict each other. The resulting dataset comprises 11,088 visual claims in 64 languages, offering a unique resource for building and evaluating misinformation-detection systems in truly multimodal and multilingual environments. We then provide comprehensive experiments using transformer-based architectures, natural language inference models, and large language models, establishing strong baselines for future research. The results show that MultiCaption is more challenging than standard NLI tasks, requiring task-specific finetuning for strong performance. Moreover, the gains from multilingual training and testing highlight the dataset's potential for building effective multilingual fact-checking pipelines without relying on machine translation.

---

### SDFLoRA: Selective Decoupled Federated LoRA for Privacy-preserving Fine-tuning with Heterogeneous Clients

Authors: Zhikang Shen, Jianrong Lu, Haiyuan Wan, Jianhai Chen
Published: 2026-01-16
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.11219v3

Federated learning (FL) for large language models (LLMs) has attracted increasing attention as a privacy-preserving approach for adapting models over distributed data, where parameter-efficient methods such as Low-Rank Adaptation (LoRA) are widely adopted to reduce communication and memory costs. However, practical deployments often exhibit rank and data heterogeneity: clients operate under different low-rank budgets and data distributions, making direct aggregation of LoRA updates biased and unstable. Existing approaches either enforce a unified rank or align heterogeneous updates into a single shared subspace, which tends to mix transferable and client-specific directions and consequently undermines personalization. Moreover, under differential privacy (DP), perturbing such structurally mixed updates injects noise into directions that should remain purely local, leading to unnecessary utility degradation. To address these issues, we propose Selective Decoupled Federated LoRA (SDFLoRA), a structure-aware LoRA framework that decouples each client update into a shared component for aggregation and a private component that preserves client-specific semantics. Only the shared component participates in subspace alignment, while the private component remains local and uncommunicated, making the training DP-compatible and stabilizing aggregation under rank heterogeneity. By injecting noise only into the aggregated shareable update, this approach avoids perturbations to local directions and improves the utility-privacy trade-off. Experiments on multiple benchmarks demonstrate that SDFLoRA outperforms federated LoRA baselines and achieves a strong utility-privacy trade-off.

---

### Model-free policy gradient for discrete-time mean-field control

Authors: Matthieu Meunier, Huyên Pham, Christoph Reisinger
Published: 2026-01-16
Categories: math.OC, cs.LG
arXiv: https://arxiv.org/abs/2601.11217v2

We study model-free policy learning for discrete-time mean-field control (MFC) problems with finite state space and compact action space. In contrast to the extensive literature on value-based methods for MFC, policy-based approaches remain largely unexplored due to the intrinsic dependence of transition kernels and rewards on the evolving population state distribution, which prevents the direct use of likelihood-ratio estimators of policy gradients from classical single-agent reinforcement learning. We introduce a novel perturbation scheme on the state-distribution flow and prove that the gradient of the resulting perturbed value function converges to the true policy gradient as the perturbation magnitude vanishes. This construction yields a fully model-free estimator based solely on simulated trajectories and an auxiliary estimate of the sensitivity of the state distribution. Building on this framework, we develop MF-REINFORCE, a model-free policy gradient algorithm for MFC, and establish explicit quantitative bounds on its bias and mean-squared error. Numerical experiments on representative mean-field control tasks demonstrate the effectiveness of the proposed approach.

---

### T$^\star$: Progressive Block Scaling for Masked Diffusion Language Models Through Trajectory Aware Reinforcement Learning

Authors: Hanchen Xia, Baoyou Chen, Yutang Ge, Guojiang Zhao, Siyu Zhu
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11214v5

We present T$^\star$, a simple TraceRL-based training curriculum for progressive block-size scaling in masked diffusion language models (MDMs). Starting from an AR-initialized small-block MDM, T$^\star$ transitions smoothly to larger blocks, enabling higher-parallelism decoding with minimal performance degradation on math reasoning benchmarks. Moreover, further analysis suggests that T$^\star$ may actually converge to an alternative decoding schedule that achieves comparable performance.

---

### VidLeaks: Membership Inference Attacks Against Text-to-Video Models

Authors: Li Wang, Wenyu Chen, Ning Yu, Zheng Li, Shanqing Guo
Published: 2026-01-16
Categories: cs.CR, cs.CV
arXiv: https://arxiv.org/abs/2601.11210v1

The proliferation of powerful Text-to-Video (T2V) models, trained on massive web-scale datasets, raises urgent concerns about copyright and privacy violations. Membership inference attacks (MIAs) provide a principled tool for auditing such risks, yet existing techniques - designed for static data like images or text - fail to capture the spatio-temporal complexities of video generation. In particular, they overlook the sparsity of memorization signals in keyframes and the instability introduced by stochastic temporal dynamics. In this paper, we conduct the first systematic study of MIAs against T2V models and introduce a novel framework VidLeaks, which probes sparse-temporal memorization through two complementary signals: 1) Spatial Reconstruction Fidelity (SRF), using a Top-K similarity to amplify spatial memorization signals from sparsely memorized keyframes, and 2) Temporal Generative Stability (TGS), which measures semantic consistency across multiple queries to capture temporal leakage. We evaluate VidLeaks under three progressively restrictive black-box settings - supervised, reference-based, and query-only. Experiments on three representative T2V models reveal severe vulnerabilities: VidLeaks achieves AUC of 82.92% on AnimateDiff and 97.01% on InstructVideo even in the strict query-only setting, posing a realistic and exploitable privacy risk. Our work provides the first concrete evidence that T2V models leak substantial membership information through both sparse and temporal memorization, establishing a foundation for auditing video generation systems and motivating the development of new defenses. Code is available at: https://zenodo.org/records/17972831.

---

### LoRA as Oracle

Authors: Marco Arazzi, Antonino Nocera
Published: 2026-01-16
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2601.11207v2

Practitioners increasingly deploy neural networks they did not train, and must audit them after the fact for hidden backdoors, without the training pipeline, the poisoned data, or knowledge of any trigger. We introduce a low-rank auditing lens built on a single observation: what a model has internalized and how it behaves are distinct axes that can diverge. Fitting a small low-rank adapter toward a hypothesis and reading the geometry of the resulting update, its energy relative to, and its alignment with, the frozen weights, measures internalization directly, independently of the model's output behavior. The lens earns its value where the two axes diverge: a backdoor is malicious internalization, a learned trigger-to-target shortcut, that behavioral auditing can miss. Reading it, we identify the backdoor's target class label-free and without any triggered data when the backdoor leaves a legible internalization signature, then erase the identified shortcut within the same low-rank subspace, uniquely coupling detection confidence to the size of the repair. Across four datasets and four architectures, our lens is the most consistent target identifier among defenses in its threat model; its rank-r repair removes backdoors while preserving clean accuracy where full-model baselines collapse it, at orders-of-magnitude lower parameter and memory cost; and it is the only method that both erases the backdoor and leaves a benign model intact, because it can decline to act when uncertain.

---

### Epistemic Control and the Normativity of Machine Learning-Based Science

Authors: Emanuele Ratti
Published: 2026-01-16
Categories: cs.CY, cs.AI
arXiv: https://arxiv.org/abs/2601.11202v1

The past few years have witnessed an increasing use of machine learning (ML) systems in science. Paul Humphreys has argued that, because of specific characteristics of ML systems, human scientists are pushed out of the loop of science. In this chapter, I investigate to what extent this is true. First, I express these concerns in terms of what I call epistemic control. I identify two conditions for epistemic control, called tracking and tracing, drawing on works in philosophy of technology. With this new understanding of the problem, I then argue against Humphreys pessimistic view. Finally, I construct a more nuanced view of epistemic control in ML-based science.

---

### FAQ: Mitigating Quantization Error via Regenerating Calibration Data with Family-Aware Quantization

Authors: Haiyang Xiao, Weiqing Li, Jinyue Guo, Guochao Jiang, Guohua Liu, Yuewei Zhang
Published: 2026-01-16
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.11200v1

Although post-training quantization (PTQ) provides an efficient numerical compression scheme for deploying large language models (LLMs) on resource-constrained devices, the representativeness and universality of calibration data remain a core bottleneck in determining the accuracy of quantization parameters. Traditional PTQ methods typically rely on limited samples, making it difficult to capture the activation distribution during the inference phase, leading to biases in quantization parameters. To address this, we propose \textbf{FAQ} (Family-Aware Quantization), a calibration data regeneration framework that leverages prior knowledge from LLMs of the same family to generate high-fidelity calibration samples. Specifically, FAQ first inputs the original calibration samples into a larger LLM from the same family as the target model, regenerating a series of high-fidelity calibration data using a highly consistent knowledge system. Subsequently, this data, carrying Chain-of-Thought reasoning and conforming to the expected activation distribution, undergoes group competition under expert guidance to select the best samples, which are then re-normalized to enhance the effectiveness of standard PTQ. Experiments on multiple model series, including Qwen3-8B, show that FAQ reduces accuracy loss by up to 28.5\% compared to the baseline with original calibration data, demonstrating its powerful potential and contribution.

---

### SD-RAG: A Prompt-Injection-Resilient Framework for Selective Disclosure in Retrieval-Augmented Generation

Authors: Aiman Al Masoud, Marco Arazzi, Antonino Nocera
Published: 2026-01-16
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2601.11199v1

Retrieval-Augmented Generation (RAG) has attracted significant attention due to its ability to combine the generative capabilities of Large Language Models (LLMs) with knowledge obtained through efficient retrieval mechanisms over large-scale data collections. Currently, the majority of existing approaches overlook the risks associated with exposing sensitive or access-controlled information directly to the generation model. Only a few approaches propose techniques to instruct the generative model to refrain from disclosing sensitive information; however, recent studies have also demonstrated that LLMs remain vulnerable to prompt injection attacks that can override intended behavioral constraints. For these reasons, we propose a novel approach to Selective Disclosure in Retrieval-Augmented Generation, called SD-RAG, which decouples the enforcement of security and privacy constraints from the generation process itself. Rather than relying on prompt-level safeguards, SD-RAG applies sanitization and disclosure controls during the retrieval phase, prior to augmenting the language model's input. Moreover, we introduce a semantic mechanism to allow the ingestion of human-readable dynamic security and privacy constraints together with an optimized graph-based data model that supports fine-grained, policy-aware retrieval. Our experimental evaluation demonstrates the superiority of SD-RAG over baseline existing approaches, achieving up to a $58\%$ improvement in the privacy score, while also showing a strong resilience to prompt injection attacks targeting the generative model.

---

### Artificial Intelligence and the US Economy: An Accounting Perspective on Investment and Production

Authors: Luisa Carpinelli, Filippo Natoli, Marco Taboga
Published: 2026-01-16
Categories: econ.GN, cs.AI
arXiv: https://arxiv.org/abs/2601.11196v1

Artificial intelligence (AI) has moved to the center of policy, market, and academic debates, but its macroeconomic footprint is still only partly understood. This paper provides an overview on how the current AI wave is captured in US national accounts, combining a simple macro-accounting framework with a stylized description of the AI production process. We highlight the crucial role played by data centers, which constitute the backbone of the AI ecosystem and have attracted formidable investment in 2025, as they are indispensable for meeting the rapidly increasing worldwide demand for AI services. We document that the boom in IT and AI-related capital expenditure in the first three quarters of the year has given an outsized boost to aggregate demand, while its contribution to GDP growth is smaller once the high import content of AI hardware is netted out. Furthermore, simple calculations suggest that, at current utilization rates and pricing, the production of services originating in new AI data centers could contribute to GDP over the turn of the next quarters on a scale comparable to that of investment spending to date. Short reinvestment cycles and uncertainty about future AI demand, while not currently acting as a macroeconomic drag, can nevertheless fuel macroeconomic risks over the medium term.

---

### ATATA: One Algorithm to Align Them All

Authors: Boyi Pang, Savva Ignatyev, Vladimir Ippolitov, Ramil Khafizov, Yurii Melnik, Oleg Voynov, Maksim Nakhodnov, Aibek Alanov, Xiaopeng Fan, Peter Wonka, Evgeny Burnaev
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11194v2

We suggest a new multi-modal algorithm for joint inference of paired structurally aligned samples with Rectified Flow models. While some existing methods propose a codependent generation process, they do not view the problem of joint generation from a structural alignment perspective. Recent work uses Score Distillation Sampling to generate aligned 3D models, but SDS is known to be time-consuming, prone to mode collapse, and often provides cartoonish results. By contrast, our suggested approach relies on the joint transport of a segment in the sample space, yielding faster computation at inference time. Our approach can be built on top of an arbitrary Rectified Flow model operating on the structured latent space. We show the applicability of our method to the domains of image, video, and 3D shape generation using state-of-the-art baselines and evaluate it against both editing-based and joint inference-based competing approaches. We demonstrate a high degree of structural alignment for the sample pairs obtained with our method and a high visual quality of the samples. Our method improves the state-of-the-art for image and video generation pipelines. For 3D generation, it is able to show comparable quality while working orders of magnitude faster.

---

### DOREMI: Optimizing Long Tail Predictions in Document-Level Relation Extraction

Authors: Laura Menotti, Stefano Marchesin, Gianmaria Silvello
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11190v1

Document-Level Relation Extraction (DocRE) presents significant challenges due to its reliance on cross-sentence context and the long-tail distribution of relation types, where many relations have scarce training examples. In this work, we introduce DOcument-level Relation Extraction optiMizing the long taIl (DOREMI), an iterative framework that enhances underrepresented relations through minimal yet targeted manual annotations. Unlike previous approaches that rely on large-scale noisy data or heuristic denoising, DOREMI actively selects the most informative examples to improve training efficiency and robustness. DOREMI can be applied to any existing DocRE model and is effective at mitigating long-tail biases, offering a scalable solution to improve generalization on rare relations.

---

### Policy-Based Deep Reinforcement Learning Hyperheuristics for Job-Shop Scheduling Problems

Authors: Sofiene Lassoued, Asrat Gobachew, Stefan Lier, Andreas Schwung
Published: 2026-01-16
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.11189v1

This paper proposes a policy-based deep reinforcement learning hyper-heuristic framework for solving the Job Shop Scheduling Problem. The hyper-heuristic agent learns to switch scheduling rules based on the system state dynamically. We extend the hyper-heuristic framework with two key mechanisms. First, action prefiltering restricts decision-making to feasible low-level actions, enabling low-level heuristics to be evaluated independently of environmental constraints and providing an unbiased assessment. Second, a commitment mechanism regulates the frequency of heuristic switching. We investigate the impact of different commitment strategies, from step-wise switching to full-episode commitment, on both training behavior and makespan. Additionally, we compare two action selection strategies at the policy level: deterministic greedy selection and stochastic sampling. Computational experiments on standard JSSP benchmarks demonstrate that the proposed approach outperforms traditional heuristics, metaheuristics, and recent neural network-based scheduling methods

---

### TimeMar: Multi-Scale Autoregressive Modeling for Unconditional Time Series Generation

Authors: Xiangyu Xu, Qingsong Zhong, Jilin Hu
Published: 2026-01-16
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.11184v1

Generative modeling offers a promising solution to data scarcity and privacy challenges in time series analysis. However, the structural complexity of time series, characterized by multi-scale temporal patterns and heterogeneous components, remains insufficiently addressed. In this work, we propose a structure-disentangled multiscale generation framework for time series. Our approach encodes sequences into discrete tokens at multiple temporal resolutions and performs autoregressive generation in a coarse-to-fine manner, thereby preserving hierarchical dependencies. To tackle structural heterogeneity, we introduce a dual-path VQ-VAE that disentangles trend and seasonal components, enabling the learning of semantically consistent latent representations. Additionally, we present a guidance-based reconstruction strategy, where coarse seasonal signals are utilized as priors to guide the reconstruction of fine-grained seasonal patterns. Experiments on six datasets show that our approach produces higher-quality time series than existing methods. Notably, our model achieves strong performance with a significantly reduced parameter count and exhibits superior capability in generating high-quality long-term sequences. Our implementation is available at https://anonymous.4open.science/r/TimeMAR-BC5B.

---

### Democratizing planetary-scale analysis: An ultra-lightweight Earth embedding database for accurate and flexible global land monitoring

Authors: Shuang Chen, Jie Wang, Shuai Yuan, Jiayang Li, Yu Xia, Yuanhong Liao, Junbo Wei, Jincheng Yuan, Xiaoqing Xu, Xiaolin Zhu, Peng Zhu, Hongsheng Zhang et al.
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11183v1

The rapid evolution of satellite-borne Earth Observation (EO) systems has revolutionized terrestrial monitoring, yielding petabyte-scale archives. However, the immense computational and storage requirements for global-scale analysis often preclude widespread use, hindering planetary-scale studies. To address these barriers, we present Embedded Seamless Data (ESD), an ultra-lightweight, 30-m global Earth embedding database spanning the 25-year period from 2000 to 2024. By transforming high-dimensional, multi-sensor observations from the Landsat series (5, 7, 8, and 9) and MODIS Terra into information-dense, quantized latent vectors, ESD distills essential geophysical and semantic features into a unified latent space. Utilizing the ESDNet architecture and Finite Scalar Quantization (FSQ), the dataset achieves a transformative ~340-fold reduction in data volume compared to raw archives. This compression allows the entire global land surface for a single year to be encapsulated within approximately 2.4 TB, enabling decadal-scale global analysis on standard local workstations. Rigorous validation demonstrates high reconstructive fidelity (MAE: 0.0130; RMSE: 0.0179; CC: 0.8543). By condensing the annual phenological cycle into 12 temporal steps, the embeddings provide inherent denoising and a semantically organized space that outperforms raw reflectance in land-cover classification, achieving 79.74% accuracy (vs. 76.92% for raw fusion). With robust few-shot learning capabilities and longitudinal consistency, ESD provides a versatile foundation for democratizing planetary-scale research and advancing next-generation geospatial artificial intelligence.

---

### TANDEM: Temporal-Aware Neural Detection for Multimodal Hate Speech

Authors: Girish A. Koushik, Helen Treharne, Diptesh Kanojia
Published: 2026-01-16
Categories: cs.AI, cs.CL, cs.MM, cs.SI
arXiv: https://arxiv.org/abs/2601.11178v3

Social media platforms are increasingly dominated by long-form multimodal content, where harmful narratives are constructed through a complex interplay of audio, visual, and textual cues. While automated systems can flag hate speech with high accuracy, they often function as "black boxes" that fail to provide the granular, interpretable evidence, such as precise timestamps and target identities, required for effective human-in-the-loop moderation. In this work, we introduce TANDEM, a unified framework that transforms audio-visual hate detection from a binary classification task into a structured reasoning problem. Our approach employs a novel tandem reinforcement learning strategy where vision-language and audio-language models optimize each other through self-constrained cross-modal context, stabilizing reasoning over extended temporal sequences without requiring dense frame-level supervision. Experiments across three benchmark datasets demonstrate that TANDEM significantly outperforms zero-shot and context-augmented baselines, achieving 0.73 F1 in target identification on HateMM (a 30% improvement over state-of-the-art) while maintaining precise temporal grounding. We further observe that while binary detection is robust, differentiating between offensive and hateful content remains challenging in multi-class settings due to inherent label ambiguity and dataset imbalance. More broadly, our findings suggest that structured, interpretable alignment is achievable even in complex multimodal settings, offering a blueprint for the next generation of transparent and actionable online safety moderation tools.

---

### The Growing Gains and Pains of Iterative Web Corpora Crawling: Insights from South Slavic CLASSLA-web 2.0 Corpora

Authors: Taja Kuzman Pungeršek, Peter Rupnik, Vít Suchomel, Nikola Ljubešić
Published: 2026-01-16
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.11170v2

Crawling national top-level domains has proven to be highly effective for collecting texts in less-resourced languages. This approach has been recently used for South Slavic languages and resulted in the largest general corpora for this language group: the CLASSLA-web 1.0 corpora. Building on this success, we established a continuous crawling infrastructure for iterative national top-level domain crawling across South Slavic and related webs. We present the first outcome of this crawling infrastructure - the CLASSLA-web 2.0 corpus collection, with substantially larger web corpora containing 17.0 billion words in 38.1 million texts in seven languages: Bosnian, Bulgarian, Croatian, Macedonian, Montenegrin, Serbian, and Slovenian. In addition to genre categories, the new version is also automatically annotated with topic labels. Comparing CLASSLA-web 2.0 with its predecessor reveals that only one-fifth of the texts overlap, showing that re-crawling after just two years yields largely new content. However, while the new web crawls bring growing gains, we also notice growing pains - a manual inspection of top domains reveals a visible degradation of web content, as machine-generated sites now contribute a significant portion of texts.

---

### SoLA-Vision: Fine-grained Layer-wise Linear Softmax Hybrid Attention

Authors: Ruibang Li, Guan Luo, Yiwei Zhang, Jin Gao, Bing Li, Weiming Hu
Published: 2026-01-16
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.11164v1

Standard softmax self-attention excels in vision tasks but incurs quadratic complexity O(N^2), limiting high-resolution deployment. Linear attention reduces the cost to O(N), yet its compressed state representations can impair modeling capacity and accuracy. We present an analytical study that contrasts linear and softmax attention for visual representation learning from a layer-stacking perspective. We further conduct systematic experiments on layer-wise hybridization patterns of linear and softmax attention. Our results show that, compared with rigid intra-block hybrid designs, fine-grained layer-wise hybridization can match or surpass performance while requiring fewer softmax layers. Building on these findings, we propose SoLA-Vision (Softmax-Linear Attention Vision), a flexible layer-wise hybrid attention backbone that enables fine-grained control over how linear and softmax attention are integrated. By strategically inserting a small number of global softmax layers, SoLA-Vision achieves a strong trade-off between accuracy and computational cost. On ImageNet-1K, SoLA-Vision outperforms purely linear and other hybrid attention models. On dense prediction tasks, it consistently surpasses strong baselines by a considerable margin. Code will be released.

---