# Research Papers — 2026-01 · bundle page 85 of 189

> Full-text markdown of papers #4201–4250 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/85
[← Previous bundle](https://aipod.dev/papers/archive/2026-01/page/84.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-01/page/86.md)

### CARPE: Context-Aware Image Representation Prioritization via Ensemble for Large Vision-Language Models

Authors: Donghee Lee, Rui Cai, Zhe Zhao
Published: 2026-01-20
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2601.13622v3

Large vision-language models (LVLMs) are typically trained using autoregressive language modeling objectives, which align visual representations with linguistic space. While effective for multimodal reasoning, this alignment can weaken vision-centric capabilities, causing LVLMs to underperform their base vision encoders on tasks such as image classification. To address this limitation, we propose Context-Aware Image Representation Prioritization via Ensemble (CARPE), a lightweight framework that integrates raw vision features with aligned LLM representations through vision-integration layers and a context-aware ensemble mechanism. This design enhances the model's ability to adaptively weight visual and textual modalities and enables the model to capture various aspects of image representations. Extensive experiments demonstrate that CARPE improves performance on both image classification and diverse vision-language benchmarks. Our results suggest that modality balancing plays a critical role in multimodal generalization by improving representation utilization within autoregressive LVLMs.

---

### CauScientist: Teaching LLMs to Respect Data for Causal Discovery

Authors: Bo Peng, Sirui Chen, Lei Xu, Chaochao Lu
Published: 2026-01-20
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2601.13614v1

Causal discovery is fundamental to scientific understanding and reliable decision-making. Existing approaches face critical limitations: purely data-driven methods suffer from statistical indistinguishability and modeling assumptions, while recent LLM-based methods either ignore statistical evidence or incorporate unverified priors that can mislead result. To this end, we propose CauScientist, a collaborative framework that synergizes LLMs as hypothesis-generating "data scientists" with probabilistic statistics as rigorous "verifiers". CauScientist employs hybrid initialization to select superior starting graphs, iteratively refines structures through LLM-proposed modifications validated by statistical criteria, and maintains error memory to guide efficient search space. Experiments demonstrate that CauScientist substantially outperforms purely data-driven baselines, achieving up to 53.8% F1 score improvement and enhancing recall from 35.0% to 100.0%. Notably, while standalone LLM performance degrades with graph complexity, CauScientist reduces structural hamming distance (SHD) by 44.0% compared to Qwen3-32B on 37-node graphs. Our project page is at https://github.com/OpenCausaLab/CauScientist.

---

### Fisher-Informed Parameterwise Aggregation for Federated Learning with Heterogeneous Data

Authors: Zhipeng Chang, Ting He, Wenrui Hao
Published: 2026-01-20
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.13608v1

Federated learning aggregates model updates from distributed clients, but standard first order methods such as FedAvg apply the same scalar weight to all parameters from each client. Under non-IID data, these uniformly weighted updates can be strongly misaligned across clients, causing client drift and degrading the global model. Here we propose Fisher-Informed Parameterwise Aggregation (FIPA), a second-order aggregation method that replaces client-level scalar weights with parameter-specific Fisher Information Matrix (FIM) weights, enabling true parameter-level scaling that captures how each client's data uniquely influences different parameters. With low-rank approximation, FIPA remains communication- and computation-efficient. Across nonlinear function regression, PDE learning, and image classification, FIPA consistently improves over averaging-based aggregation, and can be effectively combined with state-of-the-art client-side optimization algorithms to further improve image classification accuracy. These results highlight the benefits of FIPA for federated learning under heterogeneous data distributions.

---

### ChartVerse: Scaling Chart Reasoning via Reliable Programmatic Synthesis from Scratch

Authors: Zheng Liu, Honglin Lin, Chonghan Qin, Xiaoyang Wang, Xin Gao, Yu Li, Mengzhang Cai, Yun Zhu, Zhanping Zhong, Qizhi Pei, Zhuoshi Pan, Xiaoran Shang et al.
Published: 2026-01-20
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.13606v2

Chart reasoning is a critical capability for Vision Language Models (VLMs). However, the development of open-source models is severely hindered by the lack of high-quality training data. Existing datasets suffer from a dual challenge: synthetic charts are often simplistic and repetitive, while the associated QA pairs are prone to hallucinations and lack the reasoning depth required for complex tasks. To bridge this gap, we propose ChartVerse, a scalable framework designed to synthesize complex charts and reliable reasoning data from scratch. (1) To address the bottleneck of simple patterns, we first introduce Rollout Posterior Entropy (RPE), a novel metric that quantifies chart complexity. Guided by RPE, we develop complexity-aware chart coder to autonomously synthesize diverse, high-complexity charts via executable programs. (2) To guarantee reasoning rigor, we develop truth-anchored inverse QA synthesis. Diverging from standard generation, we adopt an answer-first paradigm: we extract deterministic answers directly from the source code, generate questions conditional on these anchors, and enforce strict consistency verification. To further elevate difficulty and reasoning depth, we filter samples based on model fail-rate and distill high-quality Chain-of-Thought (CoT) reasoning. We curate ChartVerse-SFT-600K and ChartVerse-RL-40K using Qwen3-VL-30B-A3B-Thinking as the teacher. Experimental results demonstrate that ChartVerse-8B achieves state-of-the-art performance, notably surpassing its teacher and rivaling the stronger Qwen3-VL-32B-Thinking. We release our code, model weights, and datasets in https://chartverse.github.io.

---

### Optimizing Parallel Schemes with Lyapunov Exponents and kNN-LLE Estimation

Authors: Mudassir Shams, Andrei Velichko, Bruno Carpentieri
Published: 2026-01-20
Categories: math.NA, cs.LG, math.DS
arXiv: https://arxiv.org/abs/2601.13604v1

Inverse parallel schemes remain indispensable tools for computing the roots of nonlinear systems, yet their dynamical behavior can be unexpectedly rich, ranging from strong contraction to oscillatory or chaotic transients depending on the choice of algorithmic parameters and initial states. A unified analytical-data-driven methodology for identifying, measuring, and reducing such instabilities in a family of uni-parametric inverse parallel solvers is presented in this study. On the theoretical side, we derive stability and bifurcation characterizations of the underlying iterative maps, identifying parameter regions associated with periodic or chaotic behavior. On the computational side, we introduce a micro-series pipeline based on kNN-driven estimation of the local largest Lyapunov exponent (LLE), applied to scalar time series derived from solver trajectories. The resulting sliding-window Lyapunov profiles provide fine-grained, real-time diagnostics of contractive or unstable phases and reveal transient behaviors not captured by coarse linearized analysis. Leveraging this correspondence, we introduce a Lyapunov-informed parameter selection strategy that identifies solver settings associated with stable behavior, particularly when the estimated LLE indicates persistent instability. Comprehensive experiments on ensembles of perturbed initial guesses demonstrate close agreement between the theoretical stability diagrams and empirical Lyapunov profiles, and show that the proposed adaptive mechanism significantly improves robustness. The study establishes micro-series Lyapunov analysis as a practical, interpretable tool for constructing self-stabilizing root-finding schemes and opens avenues for extending such diagnostics to higher-dimensional or noise-contaminated problems.

---

### A Gaussian Perspective for Distributional Discrepancy in Generative Diffusion Models

Authors: Qiang Sun, H. Vincent Poor, Wenyi Zhang
Published: 2026-01-20
Categories: cs.IT, cs.LG
arXiv: https://arxiv.org/abs/2601.13602v3

This paper introduces an analytical approach to quantifying and optimizing the distributional discrepancy in generative diffusion models. For a multivariate Gaussian source, we explicitly derive the closed-form evolution trajectory and the resulting Kullback-Leibler (KL) divergence between the distributions of the source data and the reversely sampled data. Asymptotic analysis via the Euler-Maclaurin expansion characterizes the convergence behavior of this KL divergence, extracting its dominant term as an explicit functional of the noise schedule. Minimizing this dominant term via the calculus of variations yields a noise schedule described by a tangent law, inherently determined by the source covariance spectrum. We further prove that the Gaussian source exhibits an extremal property for the KL divergence among general source distributions with a given covariance. We also utilize the analytical KL divergence as a principled metric to identify efficient time discretization strategies for pretrained diffusion models, and demonstrate via experiments over diverse datasets that the identified strategies consistently outperform established baselines, particularly under constrained function evaluation budgets.

---

### Foundations of Global Consistency Checking with Noisy LLM Oracles

Authors: Paul He, Elke Kirschbaum, Shiva Kasiviswanathan
Published: 2026-01-20
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.13600v1

Ensuring that collections of natural-language facts are globally consistent is essential for tasks such as fact-checking, summarization, and knowledge base construction. While Large Language Models (LLMs) can assess the consistency of small subsets of facts, their judgments are noisy, and pairwise checks are insufficient to guarantee global coherence. We formalize this problem and show that verifying global consistency requires exponentially many oracle queries in the worst case. To make the task practical, we propose an adaptive divide-and-conquer algorithm that identifies minimal inconsistent subsets (MUSes) of facts and optionally computes minimal repairs through hitting-sets. Our approach has low-degree polynomial query complexity. Experiments with both synthetic and real LLM oracles show that our method efficiently detects and localizes inconsistencies, offering a scalable framework for linguistic consistency verification with LLM-based evaluators.

---

### Diffusion In Diffusion: Reclaiming Global Coherence in Semi-Autoregressive Diffusion

Authors: Linrui Ma, Yufei Cui, Kai Han, Yunhe Wang
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13599v2

One of the most compelling features of global discrete diffusion language models is their global bidirectional contextual capability. However, existing block-based diffusion studies tend to introduce autoregressive priors, which, while offering benefits, can cause models to lose this global coherence at the macro level. To regain global contextual understanding while preserving the advantages of the semi-autoregressive paradigm, we propose Diffusion in Diffusion, a 'draft-then-refine' framework designed to overcome the irreversibility and myopia problems inherent in block diffusion models. Our approach first employs block diffusion to generate rapid drafts using small blocks, then refines these drafts through global bidirectional diffusion with a larger bidirectional receptive field. We utilize snapshot confidence remasking to identify the most critical tokens that require modification, and apply mix-scale training to expand the block diffusion model's global capabilities. Empirical results demonstrate that our approach sets a new benchmark for discrete diffusion models on the OpenWebText dataset. Using only 26% of the fine-tuning budget of baseline models, we reduce generative perplexity from 25.7 to 21.9, significantly narrowing the performance gap with autoregressive models.

---

### Machine learning based radiative parameterization scheme and its performance in operational reforecast experiments

Authors: Hao Jing, Sa Xiao, Haoyu Li, Huadong Xiao, Wei Xue
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13592v1

Radiation is typically the most time-consuming physical process in numerical models. One solution is to use machine learning methods to simulate the radiation process to improve computational efficiency. From an operational standpoint, this study investigates critical limitations inherent to hybrid forecasting frameworks that embed deep neural networks into numerical prediction models, with a specific focus on two fundamental bottlenecks: coupling compatibility and long-term integration stability. A residual convolutional neural network is employed to approximate the Rapid Radiative Transfer Model for General Circulation Models (RRTMG) within the global operational system of China Meteorological Administration. We adopted an offline training and online coupling approach. First, a comprehensive dataset is generated through model simulations, encompassing all atmospheric columns both with and without cloud cover. To ensure the stability of the hybrid model, the dataset is enhanced via experience replay, and additional output constraints based on physical significance are imposed. Meanwhile, a LibTorch-based coupling method is utilized, which is more suitable for real-time operational computations. The hybrid model is capable of performing ten-day integrated forecasts as required. A two-month operational reforecast experiment demonstrates that the machine learning emulator achieves accuracy comparable to that of the traditional physical scheme, while accelerating the computation speed by approximately eightfold.

---

### DSAEval: Evaluating Data Science Agents on a Wide Range of Real-World Data Science Problems

Authors: Maojun Sun, Yifei Xie, Yue Wu, Ruijian Han, Binyan Jiang, Defeng Sun, Yancheng Yuan, Jian Huang
Published: 2026-01-20
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2601.13591v3

Recent LLM-based data agents aim to automate data science tasks ranging from data analysis to deep learning. However, the open-ended nature of real-world data science problems, which often span multiple taxonomies and lack standard answers, poses a significant challenge for evaluation. To address this, we introduce DSAEval, a benchmark comprising 641 real-world data science problems grounded in 285 diverse datasets, covering both structured and unstructured data (e.g., image and text). DSAEval incorporates three distinctive features: (1) Multimodal Environment Perception, which enables agents to interpret observations from multiple modalities, including text and vision; (2) Multi-Query Interactions, which mirror the iterative and cumulative nature of real-world data science projects; and (3) Multi-Dimensional Evaluation, which provides a holistic assessment across reasoning, code, and results. We systematically evaluate 13 recent advanced agentic LLMs using DSAEval. Our results show that Claude-Sonnet-4.5 achieves the strongest overall performance, MiMo-V2-Pro and GPT-5.2 lead in duration and step efficiency, respectively, and MiMo-V2-Flash is the most cost-effective. We further demonstrate that multimodal perception consistently improves performance on vision-related tasks, with gains ranging from 2.04% to 11.30%. Overall, while current data science agents perform well on structured data and routine data analysis workflows, substantial challenges remain in unstructured domains. Finally, we offer critical insights and outline future research directions.

---

### Vulnerability of LLMs' Stated Beliefs? LLMs Belief Resistance Check Through Strategic Persuasive Conversation Interventions

Authors: Fan Huang, Haewoon Kwak, Jisun An
Published: 2026-01-20
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.13590v3

Large Language Models (LLMs) are increasingly employed in various question-answering tasks. However, recent studies showcase that LLMs are susceptible to persuasion and could adopt counterfactual beliefs. We present a systematic evaluation of LLM susceptibility to persuasion under the \emph{Source--Message--Channel--Receiver} (SMCR) communication framework. Across six mainstream Large Language Models (LLMs) and three domains (factual knowledge, medical QA, and social bias), we analyze how different persuasive strategies influence stated belief stability over multiple interaction turns. We further examine whether verbalized confidence prompting (i.e., eliciting self-reported confidence scores) affects resistance to persuasion. Results show that the smallest model (Llama 3.2-3B) exhibits extreme compliance, with 82.5\% of belief changes occurring at the first persuasive turn (average end turn of 1.1--1.4). Contrary to expectations, verbalized confidence prompting \emph{increases} vulnerability by accelerating belief erosion rather than enhancing robustness. Finally, an exploratory study of adversarial fine-tuning reveals highly model-dependent effectiveness: GPT-4o-mini achieves near-complete robustness (98.6\%), and Mistral~7B improves substantially (35.7\% $\rightarrow$ 79.3\%), but Llama models remain highly susceptible ($<$14\% RQ1) even when fine-tuned on their own failure cases. Together, these findings highlight substantial model-dependent limits of current robustness interventions and offer guidance for developing more trustworthy LLMs.

---

### Motion-to-Response Content Generation via Multi-Agent AI System with Real-Time Safety Verification

Authors: HyeYoung Lee
Published: 2026-01-20
Categories: cs.AI, cs.SD
arXiv: https://arxiv.org/abs/2601.13589v1

This paper proposes a multi-agent artificial intelligence system that generates response-oriented media content in real time based on audio-derived emotional signals. Unlike conventional speech emotion recognition studies that focus primarily on classification accuracy, our approach emphasizes the transformation of inferred emotional states into safe, age-appropriate, and controllable response content through a structured pipeline of specialized AI agents. The proposed system comprises four cooperative agents: (1) an Emotion Recognition Agent with CNN-based acoustic feature extraction, (2) a Response Policy Decision Agent for mapping emotions to response modes, (3) a Content Parameter Generation Agent for producing media control parameters, and (4) a Safety Verification Agent enforcing age-appropriateness and stimulation constraints. We introduce an explicit safety verification loop that filters generated content before output, ensuring compliance with predefined rules. Experimental results on public datasets demonstrate that the system achieves 73.2% emotion recognition accuracy, 89.4% response mode consistency, and 100% safety compliance while maintaining sub-100ms inference latency suitable for on-device deployment. The modular architecture enables interpretability and extensibility, making it applicable to child-adjacent media, therapeutic applications, and emotionally responsive smart devices.

---

### TREX: Tokenizer Regression for Optimal Data Mixture

Authors: Inho Won, Hangyeol Yoo, Minkyung Cho, Jungyeul Park, Hoyun Song, KyungTae Lim
Published: 2026-01-20
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.13588v1

Building effective tokenizers for multilingual Large Language Models (LLMs) requires careful control over language-specific data mixtures. While a tokenizer's compression performance critically affects the efficiency of LLM training and inference, existing approaches rely on heuristics or costly large-scale searches to determine optimal language ratios. We introduce Tokenizer Regression for Optimal Data MiXture (TREX), a regression-based framework that efficiently predicts the optimal data mixture for tokenizer training. TREX trains small-scale proxy tokenizers on random mixtures, gathers their compression statistics, and learns to predict compression performance from data mixtures. This learned model enables scalable mixture search before large-scale tokenizer training, mitigating the accuracy-cost trade-off in multilingual tokenizer design. Tokenizers trained with TReX's predicted mixtures outperform mixtures based on LLaMA3 and uniform distributions by up to 12% in both inand out-of-distribution compression efficiency, demonstrating strong scalability, robustness, and practical effectiveness.

---

### SCRIPTMIND: Crime Script Inference and Cognitive Evaluation for LLM-based Social Engineering Scam Detection System

Authors: Heedou Kim, Changsik Kim, Sanghwa Shin, Jaewoo Kang
Published: 2026-01-20
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.13581v1

Social engineering scams increasingly employ personalized, multi-turn deception, exposing the limits of traditional detection methods. While Large Language Models (LLMs) show promise in identifying deception, their cognitive assistance potential remains underexplored. We propose ScriptMind, an integrated framework for LLM-based scam detection that bridges automated reasoning and human cognition. It comprises three components: the Crime Script Inference Task (CSIT) for scam reasoning, the Crime Script-Aware Inference Dataset (CSID) for fine-tuning small LLMs, and the Cognitive Simulation-based Evaluation of Social Engineering Defense (CSED) for assessing real-time cognitive impact. Using 571 Korean phone scam cases, we built 22,712 structured scammer-sequence training instances. Experimental results show that the 11B small LLM fine-tuned with ScriptMind outperformed GPT-4o by 13%, achieving superior performance over commercial models in detection accuracy, false-positive reduction, scammer utterance prediction, and rationale quality. Moreover, in phone scam simulation experiments, it significantly enhanced and sustained users' suspicion levels, improving their cognitive awareness of scams. ScriptMind represents a step toward human-centered, cognitively adaptive LLMs for scam defense.

---

### Neural Organ Transplantation (NOT): Checkpoint-Based Modular Adaptation for Transformer Models

Authors: Ahmad Al-Zuraiqi
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13580v1

We introduce Neural Organ Transplantation (NOT), a modular adaptation framework that enables trained transformer layers to function as reusable transferable checkpoints for domain adaptation. Unlike conventional fine-tuning approaches that tightly couple trained parameters to specific model instances and training data, NOT extracts contiguous layer subsets ("donor organs") from pre-trained models, trains them independently on domain-specific data, and saves them as standalone checkpoint files that can be transplanted into compatible recipient models without access to the original training data. Through experiments on three decoder-only transformer architectures spanning 124M to 20B parameters (GPT-2, TinyLlama, and GPT-OSS), we demonstrate that donor transplantation substantially outperforms existing adaptation methods, achieving an order-of-magnitude improvement in perplexity over LoRA while training significantly faster. The method exhibits position dependence, with early insertion positions yielding optimal results. Cross-domain transfer at billion-parameter scale reveals unexpected regularization benefits. These findings demonstrate that transformer middle layers can support efficient modular transfer for decoder-only architectures, enabling privacy-preserving expertise sharing through checkpoint distribution. We note that this approach is currently limited to decoder-only models; preliminary experiments on encoder-based architectures show reduced effectiveness.

---

### FG-OrIU: Towards Better Forgetting via Feature-Gradient Orthogonality for Incremental Unlearning

Authors: Qian Feng, JiaHang Tu, Mintong Kang, Hanbin Zhao, Chao Zhang, Hui Qian
Published: 2026-01-20
Categories: cs.LG, cs.CV
arXiv: https://arxiv.org/abs/2601.13578v1

Incremental unlearning (IU) is critical for pre-trained models to comply with sequential data deletion requests, yet existing methods primarily suppress parameters or confuse knowledge without explicit constraints on both feature and gradient level, resulting in \textit{superficial forgetting} where residual information remains recoverable. This incomplete forgetting risks security breaches and disrupts retention balance, especially in IU scenarios. We propose FG-OrIU (\textbf{F}eature-\textbf{G}radient \textbf{Or}thogonality for \textbf{I}ncremental \textbf{U}nlearning), the first framework unifying orthogonal constraints on both features and gradients level to achieve deep forgetting, where the forgetting effect is irreversible. FG-OrIU decomposes feature spaces via Singular Value Decomposition (SVD), separating forgetting and remaining class features into distinct subspaces. It then enforces dual constraints: feature orthogonal projection on both forgetting and remaining classes, while gradient orthogonal projection prevents the reintroduction of forgotten knowledge and disruption to remaining classes during updates. Additionally, dynamic subspace adaptation merges newly forgetting subspaces and contracts remaining subspaces, ensuring a stable balance between removal and retention across sequential unlearning tasks. Extensive experiments demonstrate the effectiveness of our method.

---

### Comparing Without Saying: A Dataset and Benchmark for Implicit Comparative Opinion Mining from Same-User Reviews

Authors: Thanh-Lam T. Nguyen, Ngoc-Quang Le, Quoc-Trung Phu, Thi-Phuong Le, Ngoc-Huyen Pham, Phuong-Nguyen Nguyen, Hoang-Quynh Le
Published: 2026-01-20
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.13575v1

Existing studies on comparative opinion mining have mainly focused on explicit comparative expressions, which are uncommon in real-world reviews. This leaves implicit comparisons - here users express preferences across separate reviews - largely underexplored. We introduce SUDO, a novel dataset for implicit comparative opinion mining from same-user reviews, allowing reliable inference of user preferences even without explicit comparative cues. SUDO comprises 4,150 annotated review pairs (15,191 sentences) with a bi-level structure capturing aspect-level mentions and review-level preferences. We benchmark this task using two baseline architectures: traditional machine learning- and language model-based baselines. Experimental results show that while the latter outperforms the former, overall performance remains moderate, revealing the inherent difficulty of the task and establishing SUDO as a challenging and valuable benchmark for future research.

---

### Behavior Knowledge Merge in Reinforced Agentic Models

Authors: Xiangchi Yuan, Dachuan Shi, Chunhui Zhang, Zheyuan Liu, Shenglong Yao, Soroush Vosoughi, Wenke Lee
Published: 2026-01-20
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.13572v1

Reinforcement learning (RL) is central to post-training, particularly for agentic models that require specialized reasoning behaviors. In this setting, model merging offers a practical mechanism for integrating multiple RL-trained agents from different tasks into a single generalist model. However, existing merging methods are designed for supervised fine-tuning (SFT), and they are suboptimal to preserve task-specific capabilities on RL-trained agentic models. The root is a task-vector mismatch between RL and SFT: on-policy RL induces task vectors that are highly sparse and heterogeneous, whereas SFT-style merging implicitly assumes dense and globally comparable task vectors. When standard global averaging is applied under this mismatch, RL's non-overlapping task vectors that encode critical task-specific behaviors are reduced and parameter updates are diluted. To address this issue, we propose Reinforced Agent Merging (RAM), a distribution-aware merging framework explicitly designed for RL-trained agentic models. RAM disentangles shared and task-specific unique parameter updates, averaging shared components while selectively preserving and rescaling unique ones to counteract parameter update dilution. Experiments across multiple agent domains and model architectures demonstrate that RAM not only surpasses merging baselines, but also unlocks synergistic potential among agents to achieve performance superior to that of specialized agents in their domains.

---

### GeoDynamics: A Geometric State-Space Neural Network for Understanding Brain Dynamics on Riemannian Manifolds

Authors: Tingting Dan, Jiaqi Ding, Guorong Wu
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13570v2

State-space models (SSMs) have become a cornerstone for unraveling brain dynamics, revealing how latent neural states evolve over time and give rise to observed signals. By combining the flexibility of deep learning with the principled dynamical structure of SSMs, recent studies have achieved powerful fits to functional neuroimaging data. However, most existing approaches still view the brain as a set of loosely connected regions or impose oversimplified network priors, falling short of a truly holistic and self-organized dynamical system perspective. Brain functional connectivity (FC) at each time point naturally forms a symmetric positive definite (SPD) matrix, which resides on a curved Riemannian manifold rather than in Euclidean space. Capturing the trajectories of these SPD matrices is key to understanding how coordinated networks support cognition and behavior. To this end, we introduce GeoDynamics, a geometric state-space neural network that tracks latent brain-state trajectories directly on the high-dimensional SPD manifold. GeoDynamics embeds each connectivity matrix into a manifold-aware recurrent framework, learning smooth and geometry-respecting transitions that reveal task-driven state changes and early markers of Alzheimer's disease, Parkinson's disease, and autism. Beyond neuroscience, we validate GeoDynamics on human action recognition benchmarks (UTKinect, Florence, HDM05), demonstrating its scalability and robustness in modeling complex spatiotemporal dynamics across diverse domains.

---

### DRGW: Learning Disentangled Representations for Robust Graph Watermarking

Authors: Jiasen Li, Yanwei Liu, Zhuoyi Shang, Xiaoyan Gu, Weiping Wang
Published: 2026-01-20
Categories: cs.LG, cs.CR
arXiv: https://arxiv.org/abs/2601.13569v2

Graph-structured data is foundational to numerous web applications, and watermarking is crucial for protecting their intellectual property and ensuring data provenance. Existing watermarking methods primarily operate on graph structures or entangled graph representations, which compromise the transparency and robustness of watermarks due to the information coupling in representing graphs and uncontrollable discretization in transforming continuous numerical representations into graph structures. This motivates us to propose DRGW, the first graph watermarking framework that addresses these issues through disentangled representation learning. Specifically, we design an adversarially trained encoder that learns an invariant structural representation against diverse perturbations and derives a statistically independent watermark carrier, ensuring both robustness and transparency of watermarks. Meanwhile, we devise a graph-aware invertible neural network to provide a lossless channel for watermark embedding and extraction, guaranteeing high detectability and transparency of watermarks. Additionally, we develop a structure-aware editor that resolves the issue of latent modifications into discrete graph edits, ensuring robustness against structural perturbations. Experiments on diverse benchmark datasets demonstrate the superior effectiveness of DRGW.

---

### Self-Improvement as Coherence Optimization: A Theoretical Account

Authors: Tianyi Qiu, Ahmed Hani Ismail, Zhonghao He, Shi Feng
Published: 2026-01-20
Categories: cs.LG, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2601.13566v2

Can language models improve their accuracy without external supervision? Methods such as debate, bootstrap, and internal coherence maximization achieve this surprising feat, even matching golden finetuning performance. Yet why they work remains theoretically unclear. We show that they can all be understood as coherence optimization, the search for a context-to-behavior mapping that is most compressible and jointly predictable, with debate an exact instance and bootstrap and internal coherence maximization closely related to it. We prove that coherence optimization is equivalent to description-length regularization, and that among all such regularization schemes, coherence regularization with a prior derived from a pretrained model optimizes a lower bound of worst-case accuracy for semi-supervised learning. Our theory, supported by preliminary experiments, explains why feedback-free self-improvement works and predicts when it should succeed or fail.

---

### Learning Fine-Grained Correspondence with Cross-Perspective Perception for Open-Vocabulary 6D Object Pose Estimation

Authors: Yu Qin, Shimeng Fan, Fan Yang, Zixuan Xue, Zijie Mai, Wenrui Chen, Kailun Yang, Zhiyong Li
Published: 2026-01-20
Categories: cs.CV, cs.RO, eess.IV
arXiv: https://arxiv.org/abs/2601.13565v2

Open-vocabulary 6D object pose estimation empowers robots to manipulate arbitrary unseen objects guided solely by natural language. However, a critical limitation of existing approaches is their reliance on unconstrained global matching strategies. In open-world scenarios, trying to match anchor features against the entire query image space introduces excessive ambiguity, as target features are easily confused with background distractors. To resolve this, we propose Fine-grained Correspondence Pose Estimation (FiCoP), a framework that transitions from noise-prone global matching to spatially-constrained patch-level correspondence. To systematically eliminate background interference, FiCoP first employs an object-centric disentanglement step to isolate the target from macro-level environmental noise. Building upon this localized region, our core methodological innovations are twofold. Firstly, a Cross-Perspective Global Perception (CPGP) module is proposed to fuse dual-view features, establishing structural consensus through explicit context reasoning and text-guided semantic injection. Secondly, we design a Patch Correlation Predictor (PCP) that leverages a patch-to-patch correlation matrix as a structural prior. This generates a precise block-wise association map, acting as a spatial filter to enforce fine-grained, noise-resilient matching. Experiments on the REAL275 and Toyota-Light datasets demonstrate that FiCoP improves Average Recall by 8.0% and 6.1%, respectively, compared to the state-of-the-art method, highlighting its capability to deliver robust and generalized perception for robotic agents operating in complex, unconstrained open-world environments. The source code will be made publicly available at https://github.com/zjjqinyu/FiCoP.

---

### Multi-objective fluorescent molecule design with a data-physics dual-driven generative framework

Authors: Yanheng Li, Zhichen Pu, Lijiang Yang, Zehao Zhou, Yi Qin Gao
Published: 2026-01-20
Categories: cs.LG, cs.AI, physics.chem-ph, q-bio.BM
arXiv: https://arxiv.org/abs/2601.13564v1

Designing fluorescent small molecules with tailored optical and physicochemical properties requires navigating vast, underexplored chemical space while satisfying multiple objectives and constraints. Conventional generate-score-screen approaches become impractical under such realistic design specifications, owing to their low search efficiency, unreliable generalizability of machine-learning prediction, and the prohibitive cost of quantum chemical calculation. Here we present LUMOS, a data-and-physics driven framework for inverse design of fluorescent molecules. LUMOS couples generator and predictor within a shared latent representation, enabling direct specification-to-molecule design and efficient exploration. Moreover, LUMOS combines neural networks with a fast time-dependent density functional theory (TD-DFT) calculation workflow to build a suite of complementary predictors spanning different trade-offs in speed, accuracy, and generalizability, enabling reliable property prediction across diverse scenarios. Finally, LUMOS employs a property-guided diffusion model integrated with multi-objective evolutionary algorithms, enabling de novo design and molecular optimization under multiple objectives and constraints. Across comprehensive benchmarks, LUMOS consistently outperforms baseline models in terms of accuracy, generalizability and physical plausibility for fluorescence property prediction, and demonstrates superior performance in multi-objective scaffold- and fragment-level molecular optimization. Further validation using TD-DFT and molecular dynamics (MD) simulations demonstrates that LUMOS can generate valid fluorophores that meet various target specifications. Overall, these results establish LUMOS as a data-physics dual-driven framework for general fluorophore inverse design.

---

### ButterflyMoE: Compression-Scalable Ternary Experts via Structured Butterfly Orbits

Authors: Aryan Karmore
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13563v5

In current Mixture of Experts (MoE) architectures, linear memory scaling is present, the memory grows as the number of experts increases. $N$ independent expert weight matrices require $\mathcal{O}(N \cdot d^2)$ memory which exceeds the memory budget of edge devices. Current compression methods like quantization, pruning, and low-rank factorization reduce constant factors, but the scaling bottleneck is still unresolved. We introduce ButterflyMoE, a method which treats experts not as independent matrices but as geometric reorientations of a shared quantized substrate. Diversity amongst the experts arises from viewing different angles of the shared capacity and not from redundant storage. Learned rotations are applied to a shared ternary prototype. With this, each expert yields $\mathcal{O}(d^2 + N \cdot d \log d)$ memory-reducing per-expert cost from $\mathcal{O}(d^2)$ to $\mathcal{O}(d \log d)$. The key insight is that training these rotations with quantization reduces activation outliers and stabilizes extreme low-bit training where other static methods collapse. Across language modeling benchmarks, ButterflyMoE achieves 80$\times$ memory reduction at 8 experts with a highly favorable memory-accuracy tradeoff.At this 80x compression ButterflyMoE outperforms an equal memory dense baseline, showing that orbital parameterization extracts fundamentally more utility per byte. When scaled up to 256 experts, ButterflyMoE asymptotically compresses the memory by 150 $\times$. ButterflyMoE reduces the constant factor of linear scaling with compression ratio growing with the expert count.

---

### Reasoning is a Modality

Authors: Zhiguang Liu, Yi Shang
Published: 2026-01-20
Categories: cs.AI, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2601.13562v1

The Abstraction and Reasoning Corpus (ARC) provides a compact laboratory for studying abstract reasoning, an ability central to human intelligence. Modern AI systems, including LLMs and ViTs, largely operate as sequence-of-behavior prediction machines: they match observable behaviors by modeling token statistics without a persistent, readable mental state. This creates a gap with human-like behavior: humans can explain an action by decoding internal state, while AI systems can produce fluent post-hoc rationalizations that are not grounded in such a state. We hypothesize that reasoning is a modality: reasoning should exist as a distinct channel separate from the low-level workspace on which rules are applied. To test this hypothesis, on solving ARC tasks as a visual reasoning problem, we designed a novel role-separated transformer block that splits global controller tokens from grid workspace tokens, enabling iterative rule execution. Trained and evaluated within the VARC vision-centric protocol, our method achieved 62.6% accuracy on ARC-1, surpassing average human performance (60.2%) and outperforming prior methods significantly. Qualitatively, our models exhibit more coherent rule-application structure than the dense ViT baseline, consistent with a shift away from plausible probability blobs toward controller-driven reasoning.

---

### AgentGC: Evolutionary Learning-based Lossless Compression for Genomics Data with LLM-driven Multiple Agent

Authors: Sun Hui, Ding Yanfeng, Huidong Ma, Chang Xu, Keyan Jin, Lizheng Zu, Cheng Zhong, xiaoguang Liu, Gang Wang, Wentong Cai
Published: 2026-01-20
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.13559v1

Lossless compression has made significant advancements in Genomics Data (GD) storage, sharing and management. Current learning-based methods are non-evolvable with problems of low-level compression modeling, limited adaptability, and user-unfriendly interface. To this end, we propose AgentGC, the first evolutionary Agent-based GD Compressor, consisting of 3 layers with multi-agent named Leader and Worker. Specifically, the 1) User layer provides a user-friendly interface via Leader combined with LLM; 2) Cognitive layer, driven by the Leader, integrates LLM to consider joint optimization of algorithm-dataset-system, addressing the issues of low-level modeling and limited adaptability; and 3) Compression layer, headed by Worker, performs compression & decompression via a automated multi-knowledge learning-based compression framework. On top of AgentGC, we design 3 modes to support diverse scenarios: CP for compression-ratio priority, TP for throughput priority, and BM for balanced mode. Compared with 14 baselines on 9 datasets, the average compression ratios gains are 16.66%, 16.11%, and 16.33%, the throughput gains are 4.73x, 9.23x, and 9.15x, respectively.

---

### Leveraging ChatGPT and Other NLP Methods for Identifying Risk and Protective Behaviors in MSM: Social Media and Dating apps Text Analysis

Authors: Mehrab Beikzadeh, Chenglin Hong, Cory J Cascalheira, Callisto Boka, Majid Sarrafzadeh, Ian W Holloway
Published: 2026-01-20
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2601.13558v1

Men who have sex with men (MSM) are at elevated risk for sexually transmitted infections and harmful drinking compared to heterosexual men. Text data collected from social media and dating applications may provide new opportunities for personalized public health interventions by enabling automatic identification of risk and protective behaviors. In this study, we evaluated whether text from social media and dating apps can be used to predict sexual risk behaviors, alcohol use, and pre-exposure prophylaxis (PrEP) uptake among MSM. With participant consent, we collected textual data and trained machine learning models using features derived from ChatGPT embeddings, BERT embeddings, LIWC, and a dictionary-based risk term approach. The models achieved strong performance in predicting monthly binge drinking and having more than five sexual partners, with F1 scores of 0.78, and moderate performance in predicting PrEP use and heavy drinking, with F1 scores of 0.64 and 0.63. These findings demonstrate that social media and dating app text data can provide valuable insights into risk and protective behaviors and highlight the potential of large language model-based methods to support scalable and personalized public health interventions for MSM.

---

### DiffFace-Edit: A Diffusion-Based Facial Dataset for Forgery-Semantic Driven Deepfake Detection Analysis

Authors: Feng Ding, Wenhui Yi, Xinan He, Mengyao Xiao, Jianfeng Xu, Jianqiang Du
Published: 2026-01-20
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.13551v1

Generative models now produce imperceptible, fine-grained manipulated faces, posing significant privacy risks. However, existing AI-generated face datasets generally lack focus on samples with fine-grained regional manipulations. Furthermore, no researchers have yet studied the real impact of splice attacks, which occur between real and manipulated samples, on detectors. We refer to these as detector-evasive samples. Based on this, we introduce the DiffFace-Edit dataset, which has the following advantages: 1) It contains over two million AI-generated fake images. 2) It features edits across eight facial regions (e.g., eyes, nose) and includes a richer variety of editing combinations, such as single-region and multi-region edits. Additionally, we specifically analyze the impact of detector-evasive samples on detection models. We conduct a comprehensive analysis of the dataset and propose a cross-domain evaluation that combines IMDL methods. Dataset will be available at https://github.com/ywh1093/DiffFace-Edit.

---

### Patterning: The Dual of Interpretability

Authors: George Wang, Daniel Murfet
Published: 2026-01-20
Categories: cs.LG
arXiv: https://arxiv.org/abs/2601.13548v1

Mechanistic interpretability aims to understand how neural networks generalize beyond their training data by reverse-engineering their internal structures. We introduce patterning as the dual problem: given a desired form of generalization, determine what training data produces it. Our approach is based on susceptibilities, which measure how posterior expectation values of observables respond to infinitesimal shifts in the data distribution. Inverting this linear response relationship yields the data intervention that steers the model toward a target internal configuration. We demonstrate patterning in a small language model, showing that re-weighting training data along principal susceptibility directions can accelerate or delay the formation of structure, such as the induction circuit. In a synthetic parentheses balancing task where multiple algorithms achieve perfect training accuracy, we show that patterning can select which algorithm the model learns by targeting the local learning coefficient of each solution. These results establish that the same mathematical framework used to read internal structure can be inverted to write it.

---

### HateXScore: A Metric Suite for Evaluating Reasoning Quality in Hate Speech Explanations

Authors: Yujia Hu, Roy Ka-Wei Lee
Published: 2026-01-20
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.13547v1

Hateful speech detection is a key component of content moderation, yet current evaluation frameworks rarely assess why a text is deemed hateful. We introduce \textsf{HateXScore}, a four-component metric suite designed to evaluate the reasoning quality of model explanations. It assesses (i) conclusion explicitness, (ii) faithfulness and causal grounding of quoted spans, (iii) protected group identification (policy-configurable), and (iv) logical consistency among these elements. Evaluated on six diverse hate speech datasets, \textsf{HateXScore} is intended as a diagnostic complement to reveal interpretability failures and annotation inconsistencies that are invisible to standard metrics like Accuracy or F1. Moreover, human evaluation shows strong agreement with \textsf{HateXScore}, validating it as a practical tool for trustworthy and transparent moderation. \textcolor{red}{Disclaimer: This paper contains sensitive content that may be disturbing to some readers.}

---

### ChatAD: Reasoning-Enhanced Time-Series Anomaly Detection with Multi-Turn Instruction Evolution

Authors: Hui Sun, Chang Xu, Haonan Xie, Hao Li, Yuhao Huang, Chuheng Zhang, Ming Jin, Xiaoguang Liu, Gang Wang, Jiang Bian
Published: 2026-01-20
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.13546v1

LLM-driven Anomaly Detection (AD) helps enhance the understanding and explanatory abilities of anomalous behaviors in Time Series (TS). Existing methods face challenges of inadequate reasoning ability, deficient multi-turn dialogue capability, and narrow generalization. To this end, we 1) propose a multi-agent-based TS Evolution algorithm named TSEvol. On top of it, we 2) introduce the AD reasoning and multi-turn dialogue Dataset TSEData-20K and contribute the Chatbot family for AD, including ChatAD-Llama3-8B, Qwen2.5-7B, and Mistral-7B. Furthermore, 3) we propose the TS Kahneman-Tversky Optimization (TKTO) to enhance ChatAD's cross-task generalization capability. Lastly, 4) we propose a LLM-driven Learning-based AD Benchmark LLADBench to evaluate the performance of ChatAD and nine baselines across seven datasets and tasks. Our three ChatAD models achieve substantial gains, up to 34.50% in accuracy, 34.71% in F1, and a 37.42% reduction in false positives. Besides, via KTKO, our optimized ChatAD achieves competitive performance in reasoning and cross-task generalization on classification, forecasting, and imputation.

---

### TruthTensor: Evaluating LLMs through Human Imitation on Prediction Market under Drift and Holistic Reasoning

Authors: Shirin Shahabi, Spencer Graham, Haruna Isah
Published: 2026-01-20
Categories: cs.AI, cs.ET, cs.MA
arXiv: https://arxiv.org/abs/2601.13545v3

Evaluating language models and AI agents remains fundamentally challenging because static benchmarks fail to capture real-world uncertainty, distribution shift, and the gap between isolated task accuracy and human-aligned decision-making under evolving conditions. This paper introduces TruthTensor, a novel, reproducible evaluation paradigm that measures reasoning models not only as prediction engines but as human-imitation systems operating in socially-grounded, high-entropy environments. Building on forward-looking, contamination-free tasks, our framework anchors evaluation to live prediction markets and combines probabilistic scoring to provide a holistic view of model behavior. TruthTensor complements traditional correctness metrics with drift-centric diagnostics and explicit robustness checks for reproducibility. It specify human vs. automated evaluation roles, annotation protocols, and statistical testing procedures to ensure interpretability and replicability of results. In experiments across 500+ real markets (political, economic, cultural, technological), TruthTensor demonstrates that models with similar forecast accuracy can diverge markedly in calibration, drift, and risk-sensitivity, underscoring the need to evaluate models along multiple axes (accuracy, calibration, narrative stability, cost, and resource efficiency). TruthTensor therefore operationalizes modern evaluation best practices, clear hypothesis framing, careful metric selection, transparent compute/cost reporting, human-in-the-loop validation, and open, versioned evaluation contracts, to produce defensible assessments of LLMs in real-world decision contexts. We publicly released TruthTensor at https://truthtensor.com.

---

### Refined Gradient-Based Temperature Optimization for the Replica-Exchange Monte-Carlo Method

Authors: Tatsuya Miyata, Shunta Arai, Satoshi Takabe
Published: 2026-01-20
Categories: physics.comp-ph, cs.LG
arXiv: https://arxiv.org/abs/2601.13542v1

The replica-exchange Monte-Carlo (RXMC) method is a powerful Markov-chain Monte-Carlo algorithm for sampling from multi-modal distributions, which are challenging for conventional methods. The sampling efficiency of the RXMC method depends highly on the selection of the temperatures, and finding optimal temperatures remains a challenge. In this study, we propose a refined online temperature selection method by extending the gradient-based optimization framework proposed previously. Building upon the existing temperature update approach, we introduce a reparameterization technique to strictly enforce physical constraints, such as the monotonic ordering of inverse temperatures, which were not explicitly addressed in the original formulation. The proposed method defines the variance of acceptance rates between adjacent replicas as a loss function, estimates its gradient using differential information from the sampling process, and optimizes the temperatures via gradient descent. We demonstrate the effectiveness of our method through experiments on benchmark spin systems, including the two-dimensional ferromagnetic Ising model, the two-dimensional ferromagnetic XY model, and the three-dimensional Edwards-Anderson model. Our results show that the method successfully achieves uniform acceptance rates and reduces round-trip times across the temperature space. Furthermore, our proposed method offers a significant advantage over recently proposed policy gradient method that require careful hyperparameter tuning, while simultaneously preventing the constraint violations that destabilize optimization.

---

### When Wording Steers the Evaluation: Framing Bias in LLM judges

Authors: Yerin Hwang, Dongryeol Lee, Taegwan Kang, Minwoo Lee, Kyomin Jung
Published: 2026-01-20
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2601.13537v1

Large language models (LLMs) are known to produce varying responses depending on prompt phrasing, indicating that subtle guidance in phrasing can steer their answers. However, the impact of this framing bias on LLM-based evaluation, where models are expected to make stable and impartial judgments, remains largely underexplored. Drawing inspiration from the framing effect in psychology, we systematically investigate how deliberate prompt framing skews model judgments across four high-stakes evaluation tasks. We design symmetric prompts using predicate-positive and predicate-negative constructions and demonstrate that such framing induces significant discrepancies in model outputs. Across 14 LLM judges, we observe clear susceptibility to framing, with model families showing distinct tendencies toward agreement or rejection. These findings suggest that framing bias is a structural property of current LLM-based evaluation systems, underscoring the need for framing-aware protocols.

---

### Diff-MN: Diffusion Parameterized MoE-NCDE for Continuous Time Series Generation with Irregular Observations

Authors: Xu Zhang, Junwei Deng, Chang Xu, Hao Li, Jiang Bian
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13534v3

Time series generation (TSG) is widely used across domains, yet most existing methods assume regular sampling and fixed output resolutions. These assumptions are often violated in practice, where observations are irregular and sparse, while downstream applications require continuous and high-resolution TS. Although Neural Controlled Differential Equation (NCDE) is promising for modeling irregular TS, it is constrained by a single dynamics function, tightly coupled optimization, and limited ability to adapt learned dynamics to newly generated samples from the generative model. We propose Diff-MN, a continuous TSG framework that enhances NCDE with a Mixture-of-Experts (MoE) dynamics function and a decoupled architectural design for dynamics-focused training. To further enable NCDE to generalize to newly generated samples, Diff-MN employs a diffusion model to parameterize the NCDE temporal dynamics parameters (MoE weights), i.e., jointly learn the distribution of TS data and MoE weights. This design allows sample-specific NCDE parameters to be generated for continuous TS generation. Experiments on ten public and synthetic datasets demonstrate that Diff-MN consistently outperforms strong baselines on both irregular-to-regular and irregular-to-continuous TSG tasks. The code is available at the link https://github.com/microsoft/TimeCraft/tree/main/Diff-MN.

---

### Reasoning While Recommending: Entropy-Guided Latent Reasoning in Generative Re-ranking Models

Authors: Changshuo Zhang
Published: 2026-01-20
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.13533v1

Reinforcement learning plays a crucial role in generative re-ranking scenarios due to its exploration-exploitation capabilities, but existing generative methods mostly fail to adapt to the dynamic entropy changes in model difficulty during list generation, making it challenging to accurately capture complex preferences. Given that language models have achieved remarkable breakthroughs by integrating reasoning capabilities, we draw on this approach to introduce a latent reasoning mechanism, and experimental validation demonstrates that this mechanism effectively reduces entropy in the model's decision-making process. Based on these findings, we introduce the Entropy-Guided Latent Reasoning (EGLR) recommendation model, which has three core advantages. First, it abandons the "reason first, recommend later" paradigm to achieve "reasoning while recommending", specifically designed for the high-difficulty nature of list generation by enabling real-time reasoning during generation. Second, it implements entropy-guided variable-length reasoning using context-aware reasoning token alongside dynamic temperature adjustment, expanding exploration breadth in reasoning and boosting exploitation precision in recommending to achieve a more precisely adapted exploration-exploitation trade-off. Third, the model adopts a lightweight integration design with no complex independent modules or post-processing, enabling easy adaptation to existing models. Experimental results on two real-world datasets validate the model's effectiveness, and its notable advantage lies in being compatible with existing generative re-ranking models to enhance their performance. Further analyses also demonstrate its practical deployment value and research potential.

---

### Eliciting Harmful Capabilities by Fine-Tuning On Safeguarded Outputs

Authors: Jackson Kaunismaa, Avery Griffin, John Hughes, Christina Q. Knight, Mrinank Sharma, Erik Jones
Published: 2026-01-20
Categories: cs.CR, cs.AI, cs.CL, cs.LG, cs.SE
arXiv: https://arxiv.org/abs/2601.13528v1

Model developers implement safeguards in frontier models to prevent misuse, for example, by employing classifiers to filter dangerous outputs. In this work, we demonstrate that even robustly safeguarded models can be used to elicit harmful capabilities in open-source models through elicitation attacks. Our elicitation attacks consist of three stages: (i) constructing prompts in adjacent domains to a target harmful task that do not request dangerous information; (ii) obtaining responses to these prompts from safeguarded frontier models; (iii) fine-tuning open-source models on these prompt-output pairs. Since the requested prompts cannot be used to directly cause harm, they are not refused by frontier model safeguards. We evaluate these elicitation attacks within the domain of hazardous chemical synthesis and processing, and demonstrate that our attacks recover approximately 40% of the capability gap between the base open-source model and an unrestricted frontier model. We then show that the efficacy of elicitation attacks scales with the capability of the frontier model and the amount of generated fine-tuning data. Our work demonstrates the challenge of mitigating ecosystem level risks with output-level safeguards.

---

### GO-MLVTON: Garment Occlusion-Aware Multi-Layer Virtual Try-On with Diffusion Models

Authors: Yang Yu, Yunze Deng, Yige Zhang, Yanjie Xiao, Youkun Ou, Wenhao Hu, Mingchao Li, Bin Feng, Wenyu Liu, Dandan Zheng, Jingdong Chen
Published: 2026-01-20
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.13524v3

Existing image-based virtual try-on (VTON) methods primarily focus on single-layer or multi-garment VTON, neglecting multi-layer VTON (ML-VTON), which involves dressing multiple layers of garments onto the human body with realistic deformation and layering to generate visually plausible outcomes. The main challenge lies in accurately modeling occlusion relationships between inner and outer garments to reduce interference from redundant inner garment features. To address this, we propose GO-MLVTON, the first multi-layer VTON method, introducing the Garment Occlusion Learning module to learn occlusion relationships and the StableDiffusion-based Garment Morphing & Fitting module to deform and fit garments onto the human body, producing high-quality multi-layer try-on results. Additionally, we present the MLG dataset for this task and propose a new metric named Layered Appearance Coherence Difference (LACD) for evaluation. Extensive experiments demonstrate the state-of-the-art performance of GO-MLVTON. Project page: https://upyuyang.github.io/go-mlvton/.

---

### StoTAM: Stochastic Alternating Minimization for Tucker-Structured Tensor Sensing

Authors: Shuang Li
Published: 2026-01-20
Categories: cs.LG, math.OC
arXiv: https://arxiv.org/abs/2601.13522v1

Low-rank tensor sensing is a fundamental problem with broad applications in signal processing and machine learning. Among various tensor models, low-Tucker-rank tensors are particularly attractive for capturing multi-mode subspace structures in high-dimensional data. Existing recovery methods either operate on the full tensor variable with expensive tensor projections, or adopt factorized formulations that still rely on full-gradient computations, while most stochastic factorized approaches are restricted to tensor decomposition settings. In this work, we propose a stochastic alternating minimization algorithm that operates directly on the core tensor and factor matrices under a Tucker factorization. The proposed method avoids repeated tensor projections and enables efficient mini-batch updates on low-dimensional tensor factors. Numerical experiments on synthetic tensor sensing demonstrate that the proposed algorithm exhibits favorable convergence behavior in wall-clock time compared with representative stochastic tensor recovery baselines.

---

### Small Gradient Norm Regret for Online Convex Optimization

Authors: Wenzhi Gao, Chang He, Madeleine Udell
Published: 2026-01-20
Categories: stat.ML, cs.LG, math.OC
arXiv: https://arxiv.org/abs/2601.13519v4

This paper introduces a new problem-dependent regret measure for online convex optimization with smooth losses. The notion, which we call the $G^\star$ regret, depends on the cumulative squared gradient norm evaluated at the decision in hindsight. We show that the $G^\star$ regret strictly refines the existing $L^\star$ (small loss) regret, and that it can be arbitrarily sharper when the losses have vanishing curvature around the hindsight decision. We establish upper and lower bounds on the $G^\star$ regret and extend our results to dynamic regret and bandit settings. As a byproduct, we refine the existing convergence analysis of stochastic optimization algorithms in the interpolation regime. Some experiments validate our theoretical findings.

---

### AgenticRed: Evolving Agentic Systems for Red-Teaming

Authors: Jiayi Yuan, Jonathan Nöther, Natasha Jaques, Goran Radanović
Published: 2026-01-20
Categories: cs.AI, cs.NE
arXiv: https://arxiv.org/abs/2601.13518v3

While recent automated red-teaming methods show promise for systematically exposing model vulnerabilities, most existing approaches rely on human-specified workflows. This dependence on manually designed workflows suffers from human biases and makes exploring the broader design space expensive. We introduce AgenticRed, an automated pipeline that leverages LLMs' in-context learning to iteratively design and refine red-teaming systems without human intervention. Rather than optimizing attacker policies within predefined structures, AgenticRed treats red-teaming as a system design problem, and it autonomously evolves automated red-teaming systems using evolutionary selection and generational knowledge. Red-teaming systems designed by AgenticRed consistently outperform state-of-the-art approaches, achieving 96% attack success rate (ASR) on Llama-2-7B, 98% on Llama-3-8B and 100% on Qwen3-8B on HarmBench. Our approach generates robust, query-agnostic red-teaming systems that transfer strongly to the latest proprietary models, achieving an impressive 100% ASR on GPT-5.1, DeepSeek-R1 and DeepSeek V3.2. This work highlights evolutionary algorithms as a powerful approach to AI safety that can keep pace with rapidly evolving models.

---

### Automatic Adjustment of HPA Parameters and Attack Prevention in Kubernetes Using Random Forests

Authors: Hanlin Zhou, Huah Yong Chan, Jingfei Ni, Mengchun Wu, Qing Deng
Published: 2026-01-20
Categories: cs.CR, cs.AI, cs.DC
arXiv: https://arxiv.org/abs/2601.13515v1

In this paper, HTTP status codes are used as custom metrics within the HPA as the experimental scenario. By integrating the Random Forest classification algorithm from machine learning, attacks are assessed and predicted, dynamically adjusting the maximum pod parameter in the HPA to manage attack traffic. This approach enables the adjustment of HPA parameters using machine learning scripts in targeted attack scenarios while effectively managing attack traffic. All access from attacking IPs is redirected to honeypot pods, achieving a lower incidence of 5XX status codes through HPA pod adjustments under high load conditions. This method also ensures effective isolation of attack traffic, preventing excessive HPA expansion due to attacks. Additionally, experiments conducted under various conditions demonstrate the importance of setting appropriate thresholds for HPA adjustments.

---

### Autonomous computational catalysis through an agentic research system

Authors: Honghao Chen, Jiangjie Qiu, Yi Shen Tew, Xiaonan Wang
Published: 2026-01-20
Categories: cond-mat.mtrl-sci, cs.AI
arXiv: https://arxiv.org/abs/2601.13508v4

Autonomous agents are beginning to transform scientific research from tool-assisted workflows toward self-sustaining discovery processes. Computational catalysis provides a representative challenge, as catalyst discovery requires high-level questions to be translated into coordinated model construction, atomistic simulation, mechanistic analysis, and iterative design across multiple scales. Here we introduce CatMaster, a catalysis-native agentic research system that recasts computational catalysis as a low-barrier virtual ecosystem for autonomous research. CatMaster maintains an evolving research state and extends capabilities through self-feedback across model construction, calculation, critique and catalyst-design decisions within one extensible environment. Across progressively challenging tasks, CatMaster converts natural-language requests into concrete computational studies, from essential atomistic modelling and standard calculations to mechanism exploration and closed-loop catalyst design. It showed robust execution in representative computational-catalysis scenarios and near-leading performance across selected MatBench tasks, with phonons scenario demonstrating its modelling self-evolution capability. In the independent CO2-to-CO catalyst design case, CatMaster used iterative self-critique and evidence refinement to identify competitive B-CoN4 and NiN3B/N-NiN3B motifs. These results establish a virtual-ecosystem paradigm in which AI agents move beyond simulation execution toward end-to-end computational research, providing a foundation for autonomous discovery in catalysis and materials science.

---

### Anonpsy: A Graph-Based Framework for Structure-Preserving De-identification of Psychiatric Narratives

Authors: Kyung Ho Lim, Byung-Hoon Kim
Published: 2026-01-20
Categories: cs.CL
arXiv: https://arxiv.org/abs/2601.13503v2

Psychiatric narratives encode patient identity not only through explicit identifiers but also through idiosyncratic life events embedded in their clinical structure. Existing de-identification approaches, including PHI masking and LLM-based synthetic rewriting, operate at the text level and offer limited control over which semantic elements are preserved or altered. We introduce Anonpsy, a de-identification framework that reformulates the task as graph-guided semantic rewriting. Anonpsy (1) converts each narrative into a semantic graph encoding clinical entities, temporal anchors, and typed relations; (2) applies graph-constrained perturbations that modify identifying context while preserving clinically essential structure; and (3) regenerates text via graph-conditioned LLM generation. Evaluated on 90 clinician-authored psychiatric case narratives, Anonpsy preserves diagnostic fidelity while achieving consistently low re-identification risk under expert, semantic, and GPT-5-based evaluations. Compared with a strong LLM-only rewriting baseline, Anonpsy yields substantially lower semantic similarity and identifiability. These results demonstrate that explicit structural representations combined with constrained generation provide an effective approach to de-identification for psychiatric narratives.

---

### DIS2: Disentanglement Meets Distillation with Classwise Attention for Robust Remote Sensing Segmentation under Missing Modalities

Authors: Nhi Kieu, Kien Nguyen, Arnold Wiliem, Clinton Fookes, Sridha Sridharan
Published: 2026-01-20
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.13502v1

The efficacy of multimodal learning in remote sensing (RS) is severely undermined by missing modalities. The challenge is exacerbated by the RS highly heterogeneous data and huge scale variation. Consequently, paradigms proven effective in other domains often fail when confronted with these unique data characteristics. Conventional disentanglement learning, which relies on significant feature overlap between modalities (modality-invariant), is insufficient for this heterogeneity. Similarly, knowledge distillation becomes an ill-posed mimicry task where a student fails to focus on the necessary compensatory knowledge, leaving the semantic gap unaddressed. Our work is therefore built upon three pillars uniquely designed for RS: (1) principled missing information compensation, (2) class-specific modality contribution, and (3) multi-resolution feature importance. We propose a novel method DIS2, a new paradigm shifting from modality-shared feature dependence and untargeted imitation to active, guided missing features compensation. Its core novelty lies in a reformulated synergy between disentanglement learning and knowledge distillation, termed DLKD. Compensatory features are explicitly captured which, when fused with the features of the available modality, approximate the ideal fused representation of the full-modality case. To address the class-specific challenge, our Classwise Feature Learning Module (CFLM) adaptively learn discriminative evidence for each target depending on signal availability. Both DLKD and CFLM are supported by a hierarchical hybrid fusion (HF) structure using features across resolutions to strengthen prediction. Extensive experiments validate that our proposed approach significantly outperforms state-of-the-art methods across benchmarks.

---

### Optical Linear Systems Framework for Event Sensing and Computational Neuromorphic Imaging

Authors: Nimrod Kruger, Nicholas Owen Ralph, Gregory Cohen, Paul Hurley
Published: 2026-01-20
Categories: cs.CV
arXiv: https://arxiv.org/abs/2601.13498v1

Event vision sensors (neuromorphic cameras) output sparse, asynchronous ON/OFF events triggered by log-intensity threshold crossings, enabling microsecond-scale sensing with high dynamic range and low data bandwidth. As a nonlinear system, this event representation does not readily integrate with the linear forward models that underpin most computational imaging and optical system design. We present a physics-grounded processing pipeline that maps event streams to estimates of per-pixel log-intensity and intensity derivatives, and embeds these measurements in a dynamic linear systems model with a time-varying point spread function. This enables inverse filtering directly from event data, using frequency-domain Wiener deconvolution with a known (or parameterised) dynamic transfer function. We validate the approach in simulation for single and overlapping point sources under modulated defocus, and on real event data from a tunable-focus telescope imaging a star field, demonstrating source localisation and separability. The proposed framework provides a practical bridge between event sensing and model-based computational imaging for dynamic optical systems.

---

### Bridging the Gap Between Estimated and True Regret Towards Reliable Regret Estimation in Deep Learning based Mechanism Design

Authors: Shuyuan You, Zhiqiang Zhuang, Kewen Wang, Zhe Wang
Published: 2026-01-20
Categories: cs.GT, cs.LG, econ.GN
arXiv: https://arxiv.org/abs/2601.13489v1

Recent advances, such as RegretNet, ALGnet, RegretFormer and CITransNet, use deep learning to approximate optimal multi item auctions by relaxing incentive compatibility (IC) and measuring its violation via ex post regret. However, the true accuracy of these regret estimates remains unclear. Computing exact regret is computationally intractable, and current models rely on gradient based optimizers whose outcomes depend heavily on hyperparameter choices. Through extensive experiments, we reveal that existing methods systematically underestimate actual regret (In some models, the true regret is several hundred times larger than the reported regret), leading to overstated claims of IC and revenue. To address this issue, we derive a lower bound on regret and introduce an efficient item wise regret approximation. Building on this, we propose a guided refinement procedure that substantially improves regret estimation accuracy while reducing computational cost. Our method provides a more reliable foundation for evaluating incentive compatibility in deep learning based auction mechanisms and highlights the need to reassess prior performance claims in this area.

---

### The Hidden Toll of Social Media News: Causal Effects on Psychosocial Wellbeing

Authors: Olivia Pal, Agam Goyal, Eshwar Chandrasekharan, Koustuv Saha
Published: 2026-01-20
Categories: cs.SI, cs.AI, cs.CL, cs.CY, cs.HC
arXiv: https://arxiv.org/abs/2601.13487v1

News consumption on social media has become ubiquitous, yet how different forms of engagement shape psychosocial outcomes remains unclear. To address this gap, we leveraged a large-scale dataset of ~26M posts and ~45M comments on the BlueSky platform, and conducted a quasi-experimental study, matching 81,345 Treated users exposed to News feeds with 83,711 Control users using stratified propensity score analysis. We examined psychosocial wellbeing, in terms of affective, behavioral, and cognitive outcomes. Our findings reveal that news engagement produces systematic trade-offs: increased depression, stress, and anxiety, yet decreased loneliness and increased social interaction on the platform. Regression models reveal that News feed bookmarking is associated with greater psychosocial deterioration compared to commenting or quoting, with magnitude differences exceeding tenfold. These per-engagement effects accumulate with repeated exposure, showing significant psychosocial impacts. Our work extends theories of news effects beyond crisis-centric frameworks by demonstrating that routine consumption creates distinct psychological dynamics depending on engagement type, and bears implications for tools and interventions for mitigating the psychosocial costs of news consumption on social media.

---

### Towards Efficient and Robust Linguistic Emotion Diagnosis for Mental Health via Multi-Agent Instruction Refinement

Authors: Jian Zhang, Zhangqi Wang, Zhiyuan Wang, Weiping Fu, Yu He, Haiping Zhu, Qika Lin, Jun Liu
Published: 2026-01-20
Categories: cs.AI
arXiv: https://arxiv.org/abs/2601.13481v1

Linguistic expressions of emotions such as depression, anxiety, and trauma-related states are pervasive in clinical notes, counseling dialogues, and online mental health communities, and accurate recognition of these emotions is essential for clinical triage, risk assessment, and timely intervention. Although large language models (LLMs) have demonstrated strong generalization ability in emotion analysis tasks, their diagnostic reliability in high-stakes, context-intensive medical settings remains highly sensitive to prompt design. Moreover, existing methods face two key challenges: emotional comorbidity, in which multiple intertwined emotional states complicate prediction, and inefficient exploration of clinically relevant cues. To address these challenges, we propose APOLO (Automated Prompt Optimization for Linguistic Emotion Diagnosis), a framework that systematically explores a broader and finer-grained prompt space to improve diagnostic efficiency and robustness. APOLO formulates instruction refinement as a Partially Observable Markov Decision Process and adopts a multi-agent collaboration mechanism involving Planner, Teacher, Critic, Student, and Target roles. Within this closed-loop framework, the Planner defines an optimization trajectory, while the Teacher-Critic-Student agents iteratively refine prompts to enhance reasoning stability and effectiveness, and the Target agent determines whether to continue optimization based on performance evaluation. Experimental results show that APOLO consistently improves diagnostic accuracy and robustness across domain-specific and stratified benchmarks, demonstrating a scalable and generalizable paradigm for trustworthy LLM applications in mental healthcare.

---

### A Unified Variational Imputation Framework for Electric Vehicle Charging Data Using Retrieval-Augmented Language Model

Authors: Jinhao Li, Hao Wang
Published: 2026-01-20
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2601.13476v1

The reliability of data-driven applications in electric vehicle (EV) infrastructure, such as charging demand forecasting, hinges on the availability of complete, high-quality charging data. However, real-world EV datasets are often plagued by missing records, and existing imputation methods are ill-equipped for the complex, multimodal context of charging data, often relying on a restrictive one-model-per-station paradigm that ignores valuable inter-station correlations. To address these gaps, we develop a novel PRobabilistic variational imputation framework that leverages the power of large lAnguage models and retrIeval-augmented Memory (PRAIM). PRAIM employs a pre-trained language model to encode heterogeneous data, spanning time-series demand, calendar features, and geospatial context, into a unified, semantically rich representation. This is dynamically fortified by retrieval-augmented memory that retrieves relevant examples from the entire charging network, enabling a single, unified imputation model empowered by variational neural architecture to overcome data sparsity. Extensive experiments on four public datasets demonstrate that PRAIM significantly outperforms established baselines in both imputation accuracy and its ability to preserve the original data's statistical distribution, leading to substantial improvements in downstream forecasting performance.

---