id string | title string | abstract string | text string | label string | primary_category string | published string |
|---|---|---|---|---|---|---|
2512.23068v1 | Breaking the Memory Wall: Exact Analytical Differentiation via Tiled Operator-Space Evolution | Selective State Space Models (SSMs) achieve linear-time inference, yet their gradient-based sensitivity analysis remains bottlenecked by O(L) memory scaling during backpropagation. This memory constraint precludes genomic-scale modeling (L > 10^5) on consumer-grade hardware. We introduce Phase Gradient Flow (PGF), a fr... | Breaking the Memory Wall: Exact Analytical Differentiation via Tiled Operator-Space Evolution. Selective State Space Models (SSMs) achieve linear-time inference, yet their gradient-based sensitivity analysis remains bottlenecked by O(L) memory scaling during backpropagation. This memory constraint precludes genomic-sca... | lg_ml | cs.LG | 2025-12-28T20:27:58+00:00 |
2509.05877v2 | Uncertainty Quantification in Probabilistic Machine Learning Models: Theory, Methods, and Insights | Uncertainty Quantification (UQ) is essential in probabilistic machine learning models, particularly for assessing the reliability of predictions. In this paper, we present a systematic framework for estimating both epistemic and aleatoric uncertainty in probabilistic models. We focus on Gaussian Process Latent Variable... | Uncertainty Quantification in Probabilistic Machine Learning Models: Theory, Methods, and Insights. Uncertainty Quantification (UQ) is essential in probabilistic machine learning models, particularly for assessing the reliability of predictions. In this paper, we present a systematic framework for estimating both epist... | stat_ml | stat.ML | 2025-09-07T00:38:33+00:00 |
2512.06393v6 | Conflict-Aware Fusion: Mitigating Logic Inertia in Large Language Models via Structured Cognitive Priors | Large language models (LLMs) achieve high accuracy on many reasoning benchmarks but remain brittle under structural perturbations of rule-based systems. We introduce a diagnostic framework with four stress tests -- redundant vs. essential rule deletion, contradictory-rule injection, logic-preserving rewrites, and multi... | Conflict-Aware Fusion: Mitigating Logic Inertia in Large Language Models via Structured Cognitive Priors. Large language models (LLMs) achieve high accuracy on many reasoning benchmarks but remain brittle under structural perturbations of rule-based systems. We introduce a diagnostic framework with four stress tests --... | ai_general | cs.AI | 2025-12-06T10:49:50+00:00 |
2512.02532v1 | Laplace Approximation For Tensor Train Kernel Machines In System Identification | To address the scalability limitations of Gaussian process (GP) regression, several approximation techniques have been proposed. One such method is based on tensor networks, which utilizes an exponential number of basis functions without incurring exponential computational cost. However, extending this model to a fully... | Laplace Approximation For Tensor Train Kernel Machines In System Identification. To address the scalability limitations of Gaussian process (GP) regression, several approximation techniques have been proposed. One such method is based on tensor networks, which utilizes an exponential number of basis functions without i... | stat_ml | stat.ML | 2025-12-02T08:55:59+00:00 |
2512.17306v2 | Deep But Reliable: Advancing Multi-turn Reasoning for Thinking with Images | Recent advances in large Vision-Language Models (VLMs) have exhibited strong reasoning capabilities on complex visual tasks by thinking with images in their Chain-of-Thought (CoT), which is achieved by actively invoking tools to analyze visual inputs rather than merely perceiving them. However, existing models often st... | Deep But Reliable: Advancing Multi-turn Reasoning for Thinking with Images. Recent advances in large Vision-Language Models (VLMs) have exhibited strong reasoning capabilities on complex visual tasks by thinking with images in their Chain-of-Thought (CoT), which is achieved by actively invoking tools to analyze visual ... | cv_vision | cs.CV | 2025-12-19T07:44:43+00:00 |
2512.20148v1 | Enhancing annotations for 5D apple pose estimation through 3D Gaussian Splatting (3DGS) | Automating tasks in orchards is challenging because of the large amount of variation in the environment and occlusions. One of the challenges is apple pose estimation, where key points, such as the calyx, are often occluded. Recently developed pose estimation methods no longer rely on these key points, but still requir... | Enhancing annotations for 5D apple pose estimation through 3D Gaussian Splatting (3DGS). Automating tasks in orchards is challenging because of the large amount of variation in the environment and occlusions. One of the challenges is apple pose estimation, where key points, such as the calyx, are often occluded. Recent... | cv_vision | cs.CV | 2025-12-23T08:19:55+00:00 |
2512.20368v2 | Avoiding the Price of Adaptivity: Inference in Linear Contextual Bandits via Stability | Statistical inference in contextual bandits is challenging due to the adaptive, non-i.i.d. nature of the data. A growing body of work shows that classical least-squares inference can fail under adaptive sampling, and that valid confidence intervals for linear functionals typically require an inflation of order $\sqrt{d... | Avoiding the Price of Adaptivity: Inference in Linear Contextual Bandits via Stability. Statistical inference in contextual bandits is challenging due to the adaptive, non-i.i.d. nature of the data. A growing body of work shows that classical least-squares inference can fail under adaptive sampling, and that valid conf... | stat_ml | stat.ML | 2025-12-23T13:53:53+00:00 |
2512.18933v2 | Point What You Mean: Visually Grounded Instruction Policy | Vision-Language-Action (VLA) models align vision and language with embodied control, but their object referring ability remains limited when relying solely on text prompt, especially in cluttered or out-of-distribution (OOD) scenes. In this study, we introduce the Point-VLA, a plug-and-play policy that augments languag... | Point What You Mean: Visually Grounded Instruction Policy. Vision-Language-Action (VLA) models align vision and language with embodied control, but their object referring ability remains limited when relying solely on text prompt, especially in cluttered or out-of-distribution (OOD) scenes. In this study, we introduce ... | cv_vision | cs.CV | 2025-12-22T00:44:19+00:00 |
2602.00002v1 | Disentangled Interest Network for Out-of-Distribution CTR Prediction | Click-through rate (CTR) prediction, which estimates the probability of a user clicking on a given item, is a critical task for online information services. Existing approaches often make strong assumptions that training and test data come from the same distribution. However, the data distribution varies since user int... | Disentangled Interest Network for Out-of-Distribution CTR Prediction. Click-through rate (CTR) prediction, which estimates the probability of a user clicking on a given item, is a critical task for online information services. Existing approaches often make strong assumptions that training and test data come from the s... | ir_retrieval | cs.IR | 2025-11-14T18:20:44+00:00 |
2510.04780v1 | Kernel ridge regression under power-law data: spectrum and generalization | In this work, we investigate high-dimensional kernel ridge regression (KRR) on i.i.d. Gaussian data with anisotropic power-law covariance. This setting differs fundamentally from the classical source & capacity conditions for KRR, where power-law assumptions are typically imposed on the kernel eigen-spectrum itself. Ou... | Kernel ridge regression under power-law data: spectrum and generalization. In this work, we investigate high-dimensional kernel ridge regression (KRR) on i.i.d. Gaussian data with anisotropic power-law covariance. This setting differs fundamentally from the classical source & capacity conditions for KRR, where power-la... | stat_ml | stat.ML | 2025-10-06T12:58:35+00:00 |
2511.09722v1 | Masked Mineral Modeling: Continent-Scale Mineral Prospecting via Geospatial Infilling | Minerals play a critical role in the advanced energy technologies necessary for decarbonization, but characterizing mineral deposits hidden underground remains costly and challenging. Inspired by recent progress in generative modeling, we develop a learning method which infers the locations of minerals by masking and i... | Masked Mineral Modeling: Continent-Scale Mineral Prospecting via Geospatial Infilling. Minerals play a critical role in the advanced energy technologies necessary for decarbonization, but characterizing mineral deposits hidden underground remains costly and challenging. Inspired by recent progress in generative modelin... | stat_ml | stat.ML | 2025-11-12T20:28:40+00:00 |
2512.18814v1 | EchoMotion: Unified Human Video and Motion Generation via Dual-Modality Diffusion Transformer | Video generation models have advanced significantly, yet they still struggle to synthesize complex human movements due to the high degrees of freedom in human articulation. This limitation stems from the intrinsic constraints of pixel-only training objectives, which inherently bias models toward appearance fidelity at ... | EchoMotion: Unified Human Video and Motion Generation via Dual-Modality Diffusion Transformer. Video generation models have advanced significantly, yet they still struggle to synthesize complex human movements due to the high degrees of freedom in human articulation. This limitation stems from the intrinsic constraints... | cv_vision | cs.CV | 2025-12-21T17:08:14+00:00 |
2512.23180v3 | GaussianDWM: 3D Gaussian Driving World Model for Unified Scene Understanding and Multi-Modal Generation | Driving World Models (DWMs) have been developing rapidly with the advances of generative models. However, existing DWMs lack 3D scene understanding capabilities and can only generate content conditioned on input data, without the ability to interpret or reason about the driving environment. Moreover, current approaches... | GaussianDWM: 3D Gaussian Driving World Model for Unified Scene Understanding and Multi-Modal Generation. Driving World Models (DWMs) have been developing rapidly with the advances of generative models. However, existing DWMs lack 3D scene understanding capabilities and can only generate content conditioned on input dat... | cv_vision | cs.CV | 2025-12-29T03:40:05+00:00 |
2512.07109v1 | A Neural Affinity Framework for Abstract Reasoning: Diagnosing the Compositional Gap in Transformer Architectures via Procedural Task Taxonomy | Responding to Hodel et al.'s (2024) call for a formal definition of task relatedness in re-arc, we present the first 9-category taxonomy of all 400 tasks, validated at 97.5% accuracy via rule-based code analysis. We prove the taxonomy's visual coherence by training a CNN on raw grid pixels (95.24% accuracy on S3, 36.25... | A Neural Affinity Framework for Abstract Reasoning: Diagnosing the Compositional Gap in Transformer Architectures via Procedural Task Taxonomy. Responding to Hodel et al.'s (2024) call for a formal definition of task relatedness in re-arc, we present the first 9-category taxonomy of all 400 tasks, validated at 97.5% ac... | ai_general | cs.AI | 2025-12-08T02:46:00+00:00 |
2512.21999v1 | Look Closer! An Adversarial Parametric Editing Framework for Hallucination Mitigation in VLMs | While Vision-Language Models (VLMs) have garnered increasing attention in the AI community due to their promising practical applications, they exhibit persistent hallucination issues, generating outputs misaligned with visual inputs. Recent studies attribute these hallucinations to VLMs' over-reliance on linguistic pri... | Look Closer! An Adversarial Parametric Editing Framework for Hallucination Mitigation in VLMs. While Vision-Language Models (VLMs) have garnered increasing attention in the AI community due to their promising practical applications, they exhibit persistent hallucination issues, generating outputs misaligned with visual... | cv_vision | cs.CV | 2025-12-26T11:56:45+00:00 |
2512.23709v2 | Stream-DiffVSR: Low-Latency Streamable Video Super-Resolution via Auto-Regressive Diffusion | Diffusion-based video super-resolution (VSR) methods deliver strong perceptual quality but are often unsuitable for latency-sensitive scenarios due to reliance on future frames and expensive multi-step denoising. We propose Stream-DiffVSR, a causally conditioned diffusion framework for efficient online VSR. Operating s... | Stream-DiffVSR: Low-Latency Streamable Video Super-Resolution via Auto-Regressive Diffusion. Diffusion-based video super-resolution (VSR) methods deliver strong perceptual quality but are often unsuitable for latency-sensitive scenarios due to reliance on future frames and expensive multi-step denoising. We propose Str... | cv_vision | cs.CV | 2025-12-29T18:59:57+00:00 |
2511.03050v1 | Precise asymptotic analysis of Sobolev training for random feature models | Gradient information is widely useful and available in applications, and is therefore natural to include in the training of neural networks. Yet little is known theoretically about the impact of Sobolev training -- regression with both function and gradient data -- on the generalization error of highly overparameterize... | Precise asymptotic analysis of Sobolev training for random feature models. Gradient information is widely useful and available in applications, and is therefore natural to include in the training of neural networks. Yet little is known theoretically about the impact of Sobolev training -- regression with both function ... | stat_ml | stat.ML | 2025-11-04T22:49:33+00:00 |
2511.14765v1 | Optimizing Agricultural Research: A RAG-Based Approach to Mycorrhizal Fungi Information | Retrieval-Augmented Generation (RAG) represents a transformative approach within natural language processing (NLP), combining neural information retrieval with generative language modeling to enhance both contextual accuracy and factual reliability of responses. Unlike conventional Large Language Models (LLMs), which a... | Optimizing Agricultural Research: A RAG-Based Approach to Mycorrhizal Fungi Information. Retrieval-Augmented Generation (RAG) represents a transformative approach within natural language processing (NLP), combining neural information retrieval with generative language modeling to enhance both contextual accuracy and fa... | ir_retrieval | cs.IR | 2025-09-16T20:21:55+00:00 |
2512.21409v2 | kooplearn: A Scikit-Learn Compatible Library of Algorithms for Evolution Operator Learning | kooplearn is a machine-learning library that implements linear, kernel, and deep-learning estimators of dynamical operators and their spectral decompositions. kooplearn can model both discrete-time evolution operators (Koopman/Transfer) and continuous-time infinitesimal generators. By learning these operators, users ca... | kooplearn: A Scikit-Learn Compatible Library of Algorithms for Evolution Operator Learning. kooplearn is a machine-learning library that implements linear, kernel, and deep-learning estimators of dynamical operators and their spectral decompositions. kooplearn can model both discrete-time evolution operators (Koopman/T... | lg_ml | cs.LG | 2025-12-24T20:15:41+00:00 |
2512.20278v1 | Synthesizing Procedural Memory: Challenges and Architectures in Automated Workflow Generation | While CodeMem establishes executable code as the optimal representation for agentic procedural memory, the mechanism for autonomously synthesizing this memory from a blank slate remains underexplored. This paper operationalizes the transition of Large Language Models from passive tool-users to active workflow architect... | Synthesizing Procedural Memory: Challenges and Architectures in Automated Workflow Generation. While CodeMem establishes executable code as the optimal representation for agentic procedural memory, the mechanism for autonomously synthesizing this memory from a blank slate remains underexplored. This paper operationaliz... | ai_general | cs.AI | 2025-12-23T11:33:32+00:00 |
2512.21083v1 | Hierarchical Modeling Approach to Fast and Accurate Table Recognition | The extraction and use of diverse knowledge from numerous documents is a pressing challenge in intelligent information retrieval. Documents contain elements that require different recognition methods. Table recognition typically consists of three subtasks, namely table structure, cell position and cell content recognit... | Hierarchical Modeling Approach to Fast and Accurate Table Recognition. The extraction and use of diverse knowledge from numerous documents is a pressing challenge in intelligent information retrieval. Documents contain elements that require different recognition methods. Table recognition typically consists of three su... | cv_vision | cs.CV | 2025-12-24T09:58:30+00:00 |
2512.10336v1 | Multilingual VLM Training: Adapting an English-Trained VLM to French | Artificial intelligence has made great progress in recent years, particularly in the development of Vision--Language Models (VLMs) that understand both visual and textual data. However, these advancements remain largely limited to English, reducing their accessibility for non--English speakers. It is essential to exten... | Multilingual VLM Training: Adapting an English-Trained VLM to French. Artificial intelligence has made great progress in recent years, particularly in the development of Vision--Language Models (VLMs) that understand both visual and textual data. However, these advancements remain largely limited to English, reducing t... | cl_nlp | cs.CL | 2025-12-11T06:38:51+00:00 |
2512.19155v1 | Can We Test Consciousness Theories on AI? Ablations, Markers, and Robustness | The search for reliable indicators of consciousness has fragmented into competing theoretical camps (Global Workspace Theory (GWT), Integrated Information Theory (IIT), and Higher-Order Theories (HOT)), each proposing distinct neural signatures. We adopt a synthetic neuro-phenomenology approach: constructing artificial... | Can We Test Consciousness Theories on AI? Ablations, Markers, and Robustness. The search for reliable indicators of consciousness has fragmented into competing theoretical camps (Global Workspace Theory (GWT), Integrated Information Theory (IIT), and Higher-Order Theories (HOT)), each proposing distinct neural signatur... | ai_general | cs.AI | 2025-12-22T08:52:07+00:00 |
2512.24260v2 | Physically-Grounded Manifold Projection Model for Generalizable Metal Artifact Reduction in Dental CBCT | Metal artifacts in Dental CBCT severely obscure anatomical structures, hindering diagnosis. Current deep learning for Metal Artifact Reduction (MAR) faces limitations: supervised methods suffer from spectral blurring due to "regression-to-the-mean", while unsupervised ones risk structural hallucinations. Denoising Diff... | Physically-Grounded Manifold Projection Model for Generalizable Metal Artifact Reduction in Dental CBCT. Metal artifacts in Dental CBCT severely obscure anatomical structures, hindering diagnosis. Current deep learning for Metal Artifact Reduction (MAR) faces limitations: supervised methods suffer from spectral blurrin... | cv_vision | cs.CV | 2025-12-30T14:36:26+00:00 |
2509.01629v3 | Lipschitz-Guided Design of Interpolation Schedules in Generative Models | We study the design of interpolation schedules in flow and diffusion-based generative models from both statistical and numerical perspectives. Within the stochastic interpolants framework, we first show that scalar interpolation schedules are statistically equivalent under the Kullback--Leibler divergence in path space... | Lipschitz-Guided Design of Interpolation Schedules in Generative Models. We study the design of interpolation schedules in flow and diffusion-based generative models from both statistical and numerical perspectives. Within the stochastic interpolants framework, we first show that scalar interpolation schedules are stat... | stat_ml | stat.ML | 2025-09-01T17:16:34+00:00 |
2511.20177v1 | Enhancing Sequential Recommendation with World Knowledge from Large Language Models | Sequential Recommendation System~(SRS) has become pivotal in modern society, which predicts subsequent actions based on the user's historical behavior. However, traditional collaborative filtering-based sequential recommendation models often lead to suboptimal performance due to the limited information of their collabo... | Enhancing Sequential Recommendation with World Knowledge from Large Language Models. Sequential Recommendation System~(SRS) has become pivotal in modern society, which predicts subsequent actions based on the user's historical behavior. However, traditional collaborative filtering-based sequential recommendation models... | ir_retrieval | cs.IR | 2025-11-25T10:59:38+00:00 |
2512.07454v1 | Persian-Phi: Efficient Cross-Lingual Adaptation of Compact LLMs via Curriculum Learning | The democratization of AI is currently hindered by the immense computational costs required to train Large Language Models (LLMs) for low-resource languages. This paper presents Persian-Phi, a 3.8B parameter model that challenges the assumption that robust multilingual capabilities require massive model sizes or multil... | Persian-Phi: Efficient Cross-Lingual Adaptation of Compact LLMs via Curriculum Learning. The democratization of AI is currently hindered by the immense computational costs required to train Large Language Models (LLMs) for low-resource languages. This paper presents Persian-Phi, a 3.8B parameter model that challenges t... | cl_nlp | cs.CL | 2025-12-08T11:27:52+00:00 |
2511.11172v1 | Enhancing Group Recommendation using Soft Impute Singular Value Decomposition | The growing popularity of group activities increased the need to develop methods for providing recommendations to a group of users based on the collective preferences of the group members. Several group recommender systems have been proposed, but these methods often struggle due to sparsity and high-dimensionality of t... | Enhancing Group Recommendation using Soft Impute Singular Value Decomposition. The growing popularity of group activities increased the need to develop methods for providing recommendations to a group of users based on the collective preferences of the group members. Several group recommender systems have been proposed... | ir_retrieval | cs.IR | 2025-11-14T11:13:56+00:00 |
2512.22294v1 | A Three-Level Alignment Framework for Large-Scale 3D Retrieval and Controlled 4D Generation | We introduce Uni4D, a unified framework for large scale open vocabulary 3D retrieval and controlled 4D generation based on structured three level alignment across text, 3D models, and image modalities. Built upon the Align3D 130 dataset, Uni4D employs a 3D text multi head attention and search model to optimize text to ... | A Three-Level Alignment Framework for Large-Scale 3D Retrieval and Controlled 4D Generation. We introduce Uni4D, a unified framework for large scale open vocabulary 3D retrieval and controlled 4D generation based on structured three level alignment across text, 3D models, and image modalities. Built upon the Align3D 13... | cv_vision | cs.CV | 2025-12-25T20:27:24+00:00 |
2510.05952v1 | How public datasets constrain the development of diversity-aware news recommender systems, and what law could do about it | News recommender systems increasingly determine what news individuals see online. Over the past decade, researchers have extensively critiqued recommender systems that prioritise news based on user engagement. To offer an alternative, researchers have analysed how recommender systems could support the media's ability t... | How public datasets constrain the development of diversity-aware news recommender systems, and what law could do about it. News recommender systems increasingly determine what news individuals see online. Over the past decade, researchers have extensively critiqued recommender systems that prioritise news based on user... | ir_retrieval | cs.IR | 2025-10-07T14:08:38+00:00 |
2509.01306v2 | Re3: Learning to Balance Relevance & Recency for Temporal Information Retrieval | Temporal Information Retrieval (TIR) is a critical yet unresolved task for modern search systems, retrieving documents that not only satisfy a query's information need but also adhere to its temporal constraints. This task is shaped by two challenges: Relevance, ensuring alignment with the query's explicit temporal req... | Re3: Learning to Balance Relevance & Recency for Temporal Information Retrieval. Temporal Information Retrieval (TIR) is a critical yet unresolved task for modern search systems, retrieving documents that not only satisfy a query's information need but also adhere to its temporal constraints. This task is shaped by two... | ir_retrieval | cs.IR | 2025-09-01T09:44:01+00:00 |
2601.11567v1 | Measuring Stability Beyond Accuracy in Small Open-Source Medical Large Language Models for Pediatric Endocrinology | Small open-source medical large language models (LLMs) offer promising opportunities for low-resource deployment and broader accessibility. However, their evaluation is often limited to accuracy on medical multiple choice question (MCQ) benchmarks, and lacks evaluation of consistency, robustness, or reasoning behavior.... | Measuring Stability Beyond Accuracy in Small Open-Source Medical Large Language Models for Pediatric Endocrinology. Small open-source medical large language models (LLMs) offer promising opportunities for low-resource deployment and broader accessibility. However, their evaluation is often limited to accuracy on medica... | cl_nlp | cs.CL | 2025-12-26T14:30:53+00:00 |
2512.20761v3 | TS-Arena -- A Live Forecast Pre-Registration Platform | Time Series Foundation Models (TSFMs) are transforming the field of forecasting. However, evaluating them on historical data is increasingly difficult due to the risks of train-test sample overlaps and temporal overlaps between correlated train and test time series. To address this, we introduce TS-Arena, a live foreca... | TS-Arena -- A Live Forecast Pre-Registration Platform. Time Series Foundation Models (TSFMs) are transforming the field of forecasting. However, evaluating them on historical data is increasingly difficult due to the risks of train-test sample overlaps and temporal overlaps between correlated train and test time series... | lg_ml | cs.LG | 2025-12-23T20:48:11+00:00 |
2512.23340v2 | The Law of Multi-Model Collaboration: Scaling Limits of Model Ensembling for Large Language Models | Recent advances in large language models (LLMs) have been largely driven by scaling laws for individual models, which predict performance improvements as model parameters and data volume increase. However, the capabilities of any single LLM are inherently bounded. One solution originates from intricate interactions amo... | The Law of Multi-Model Collaboration: Scaling Limits of Model Ensembling for Large Language Models. Recent advances in large language models (LLMs) have been largely driven by scaling laws for individual models, which predict performance improvements as model parameters and data volume increase. However, the capabiliti... | lg_ml | cs.LG | 2025-12-29T09:55:12+00:00 |
2512.05356v2 | AI & Human Co-Improvement for Safer Co-Superintelligence | Self-improvement is a goal currently exciting the field of AI, but is fraught with danger, and may take time to fully achieve. We advocate that a more achievable and better goal for humanity is to maximize co-improvement: collaboration between human researchers and AIs to achieve co-superintelligence. That is, specific... | AI & Human Co-Improvement for Safer Co-Superintelligence. Self-improvement is a goal currently exciting the field of AI, but is fraught with danger, and may take time to fully achieve. We advocate that a more achievable and better goal for humanity is to maximize co-improvement: collaboration between human researchers ... | ai_general | cs.AI | 2025-12-05T01:50:23+00:00 |
2512.15044v1 | Agentic AI for Integrated Sensing and Communication: Analysis, Framework, and Case Study | Integrated sensing and communication (ISAC) has emerged as a key development direction in the sixth-generation (6G) era, which provides essential support for the collaborative sensing and communication of future intelligent networks. However, as wireless environments become increasingly dynamic and complex, ISAC system... | Agentic AI for Integrated Sensing and Communication: Analysis, Framework, and Case Study. Integrated sensing and communication (ISAC) has emerged as a key development direction in the sixth-generation (6G) era, which provides essential support for the collaborative sensing and communication of future intelligent networ... | ai_general | cs.AI | 2025-12-17T03:16:06+00:00 |
2512.22455v2 | AFA-LoRA: Enabling Non-Linear Adaptations in LoRA with Activation Function Annealing | Low-Rank Adaptation (LoRA) is a widely adopted parameter-efficient fine-tuning (PEFT) method. However, its linear adaptation process limits its expressive power. This means there is a gap between the expressive power of linear training and non-linear training. To bridge this gap, we propose AFA-LoRA, a novel training s... | AFA-LoRA: Enabling Non-Linear Adaptations in LoRA with Activation Function Annealing. Low-Rank Adaptation (LoRA) is a widely adopted parameter-efficient fine-tuning (PEFT) method. However, its linear adaptation process limits its expressive power. This means there is a gap between the expressive power of linear trainin... | lg_ml | cs.LG | 2025-12-27T04:12:40+00:00 |
2510.24056v2 | Copula-Stein Discrepancy: A Generator-Based Stein Operator for Archimedean Dependence | Kernel Stein discrepancies (KSDs) are widely used for goodness-of-fit testing, but standard KSDs can be insensitive to higher-order dependence features such as tail dependence. We introduce the Copula-Stein Discrepancy (CSD), which defines a Stein operator directly on the copula density to target dependence geometry ra... | Copula-Stein Discrepancy: A Generator-Based Stein Operator for Archimedean Dependence. Kernel Stein discrepancies (KSDs) are widely used for goodness-of-fit testing, but standard KSDs can be insensitive to higher-order dependence features such as tail dependence. We introduce the Copula-Stein Discrepancy (CSD), which d... | stat_ml | stat.ML | 2025-10-28T04:33:57+00:00 |
2512.07525v1 | Beyond Real: Imaginary Extension of Rotary Position Embeddings for Long-Context LLMs | Rotary Position Embeddings (RoPE) have become a standard for encoding sequence order in Large Language Models (LLMs) by applying rotations to query and key vectors in the complex plane. Standard implementations, however, utilize only the real component of the complex-valued dot product for attention score calculation. ... | Beyond Real: Imaginary Extension of Rotary Position Embeddings for Long-Context LLMs. Rotary Position Embeddings (RoPE) have become a standard for encoding sequence order in Large Language Models (LLMs) by applying rotations to query and key vectors in the complex plane. Standard implementations, however, utilize only ... | cl_nlp | cs.CL | 2025-12-08T12:59:54+00:00 |
2601.00850v1 | EdgeJury: Cross-Reviewed Small-Model Ensembles for Truthful Question Answering on Serverless Edge Inference | Hallucinations hinder reliable question answering, especially in resource-constrained deployments where frontier-scale models or retrieval pipelines may be impractical. We present EdgeJury, a lightweight ensemble framework that improves truthfulness and robustness using only small instruction-tuned language models (3B-... | EdgeJury: Cross-Reviewed Small-Model Ensembles for Truthful Question Answering on Serverless Edge Inference. Hallucinations hinder reliable question answering, especially in resource-constrained deployments where frontier-scale models or retrieval pipelines may be impractical. We present EdgeJury, a lightweight ensembl... | lg_ml | cs.LG | 2025-12-29T14:48:40+00:00 |
2509.24710v1 | MAD: Manifold Attracted Diffusion | Score-based diffusion models are a highly effective method for generating samples from a distribution of images. We consider scenarios where the training data comes from a noisy version of the target distribution, and present an efficiently implementable modification of the inference procedure to generate noiseless sam... | MAD: Manifold Attracted Diffusion. Score-based diffusion models are a highly effective method for generating samples from a distribution of images. We consider scenarios where the training data comes from a noisy version of the target distribution, and present an efficiently implementable modification of the inference ... | stat_ml | stat.ML | 2025-09-29T12:40:20+00:00 |
2512.24008v2 | SPARK: Search Personalization via Agent-Driven Retrieval and Knowledge-sharing | Personalized search demands the ability to model users' evolving, multi-dimensional information needs; a challenge for systems constrained by static profiles or monolithic retrieval pipelines. We present SPARK (Search Personalization via Agent-Driven Retrieval and Knowledge-sharing), a framework in which coordinated pe... | SPARK: Search Personalization via Agent-Driven Retrieval and Knowledge-sharing. Personalized search demands the ability to model users' evolving, multi-dimensional information needs; a challenge for systems constrained by static profiles or monolithic retrieval pipelines. We present SPARK (Search Personalization via Ag... | ai_general | cs.AI | 2025-12-30T06:09:12+00:00 |
2510.21431v1 | Oracle-Efficient Combinatorial Semi-Bandits | We study the combinatorial semi-bandit problem where an agent selects a subset of base arms and receives individual feedback. While this generalizes the classical multi-armed bandit and has broad applicability, its scalability is limited by the high cost of combinatorial optimization, requiring oracle queries at every ... | Oracle-Efficient Combinatorial Semi-Bandits. We study the combinatorial semi-bandit problem where an agent selects a subset of base arms and receives individual feedback. While this generalizes the classical multi-armed bandit and has broad applicability, its scalability is limited by the high cost of combinatorial opt... | stat_ml | stat.ML | 2025-10-24T13:07:08+00:00 |
2509.03378v9 | Understanding and Improving Shampoo and SOAP via Kullback-Leibler Minimization | Shampoo and its efficient variant, SOAP, employ structured second-moment estimations and have shown strong performance for training neural networks (NNs). In practice, however, Shampoo typically requires step-size grafting with Adam to be competitive, and SOAP mitigates this by applying Adam in Shampoo's eigenbasis -- ... | Understanding and Improving Shampoo and SOAP via Kullback-Leibler Minimization. Shampoo and its efficient variant, SOAP, employ structured second-moment estimations and have shown strong performance for training neural networks (NNs). In practice, however, Shampoo typically requires step-size grafting with Adam to be c... | stat_ml | stat.ML | 2025-09-03T14:55:15+00:00 |
2512.24772v1 | Uncertainty-aware Semi-supervised Ensemble Teacher Framework for Multilingual Depression Detection | Detecting depression from social media text is still a challenging task. This is due to different language styles, informal expression, and the lack of annotated data in many languages. To tackle these issues, we propose, Semi-SMDNet, a strong Semi-Supervised Multilingual Depression detection Network. It combines teach... | Uncertainty-aware Semi-supervised Ensemble Teacher Framework for Multilingual Depression Detection. Detecting depression from social media text is still a challenging task. This is due to different language styles, informal expression, and the lack of annotated data in many languages. To tackle these issues, we propose... | cl_nlp | cs.CL | 2025-12-31T10:35:59+00:00 |
2512.14239v1 | Two CFG Nahuatl for automatic corpora expansion | The aim of this article is to introduce two Context-Free Grammars (CFG) for Nawatl Corpora expansion. Nawatl is an Amerindian language (it is a National Language of Mexico) of the $π$-language type, i.e. a language with few digital resources. For this reason the corpora available for the learning of Large Language Mode... | Two CFG Nahuatl for automatic corpora expansion. The aim of this article is to introduce two Context-Free Grammars (CFG) for Nawatl Corpora expansion. Nawatl is an Amerindian language (it is a National Language of Mexico) of the $π$-language type, i.e. a language with few digital resources. For this reason the corpora ... | cl_nlp | cs.CL | 2025-12-16T09:49:31+00:00 |
2512.05456v1 | Do We Really Even Need Data? A Modern Look at Drawing Inference with Predicted Data | As artificial intelligence and machine learning tools become more accessible, and scientists face new obstacles to data collection (e.g., rising costs, declining survey response rates), researchers increasingly use predictions from pre-trained algorithms as substitutes for missing or unobserved data. Though appealing f... | Do We Really Even Need Data? A Modern Look at Drawing Inference with Predicted Data. As artificial intelligence and machine learning tools become more accessible, and scientists face new obstacles to data collection (e.g., rising costs, declining survey response rates), researchers increasingly use predictions from pre... | stat_ml | stat.ML | 2025-12-05T06:24:23+00:00 |
2509.14221v2 | GEM-Bench: A Benchmark for Ad-Injected Response Generation within Generative Engine Marketing | Generative Engine Marketing (GEM) is an emerging ecosystem for monetizing generative engines, such as LLM-based chatbots, by seamlessly integrating relevant advertisements into their responses. At the core of GEM lies the generation and evaluation of ad-injected responses. However, existing benchmarks are not specifica... | GEM-Bench: A Benchmark for Ad-Injected Response Generation within Generative Engine Marketing. Generative Engine Marketing (GEM) is an emerging ecosystem for monetizing generative engines, such as LLM-based chatbots, by seamlessly integrating relevant advertisements into their responses. At the core of GEM lies the gen... | ir_retrieval | cs.IR | 2025-09-17T17:53:43+00:00 |
2512.10671v1 | AEBNAS: Strengthening Exit Branches in Early-Exit Networks through Hardware-Aware Neural Architecture Search | Early-exit networks are effective solutions for reducing the overall energy consumption and latency of deep learning models by adjusting computation based on the complexity of input data. By incorporating intermediate exit branches into the architecture, they provide less computation for simpler samples, which is parti... | AEBNAS: Strengthening Exit Branches in Early-Exit Networks through Hardware-Aware Neural Architecture Search. Early-exit networks are effective solutions for reducing the overall energy consumption and latency of deep learning models by adjusting computation based on the complexity of input data. By incorporating inter... | ai_general | cs.AI | 2025-12-11T14:17:49+00:00 |
2512.12550v1 | Iterative Sampling Methods for Sinkhorn Distributionally Robust Optimization | Distributionally robust optimization (DRO) has emerged as a powerful paradigm for reliable decision-making under uncertainty. This paper focuses on DRO with ambiguity sets defined via the Sinkhorn discrepancy: an entropy-regularized Wasserstein distance, referred to as Sinkhorn DRO. Existing work primarily addresses Si... | Iterative Sampling Methods for Sinkhorn Distributionally Robust Optimization. Distributionally robust optimization (DRO) has emerged as a powerful paradigm for reliable decision-making under uncertainty. This paper focuses on DRO with ambiguity sets defined via the Sinkhorn discrepancy: an entropy-regularized Wasserste... | stat_ml | stat.ML | 2025-12-14T04:42:51+00:00 |
2512.20908v2 | Where Did This Sentence Come From? Tracing Provenance in LLM Reasoning Distillation | Reasoning distillation has attracted increasing attention. It typically leverages a large teacher model to generate reasoning paths, which are then used to fine-tune a student model so that it mimics the teacher's behavior in training contexts. However, previous approaches have lacked a detailed analysis of the origins... | Where Did This Sentence Come From? Tracing Provenance in LLM Reasoning Distillation. Reasoning distillation has attracted increasing attention. It typically leverages a large teacher model to generate reasoning paths, which are then used to fine-tune a student model so that it mimics the teacher's behavior in training ... | cl_nlp | cs.CL | 2025-12-24T03:19:05+00:00 |
2512.20714v1 | From Pilots to Practices: A Scoping Review of GenAI-Enabled Personalization in Computer Science Education | Generative AI enables personalized computer science education at scale, yet questions remain about whether such personalization supports or undermines learning. This scoping review synthesizes 32 studies (2023-2025) purposively sampled from 259 records to map personalization mechanisms and effectiveness signals in high... | From Pilots to Practices: A Scoping Review of GenAI-Enabled Personalization in Computer Science Education. Generative AI enables personalized computer science education at scale, yet questions remain about whether such personalization supports or undermines learning. This scoping review synthesizes 32 studies (2023-202... | ai_general | cs.AI | 2025-12-23T19:20:34+00:00 |
2512.12686v1 | Memoria: A Scalable Agentic Memory Framework for Personalized Conversational AI | Agentic memory is emerging as a key enabler for large language models (LLM) to maintain continuity, personalization, and long-term context in extended user interactions, critical capabilities for deploying LLMs as truly interactive and adaptive agents. Agentic memory refers to the memory that provides an LLM with agent... | Memoria: A Scalable Agentic Memory Framework for Personalized Conversational AI. Agentic memory is emerging as a key enabler for large language models (LLM) to maintain continuity, personalization, and long-term context in extended user interactions, critical capabilities for deploying LLMs as truly interactive and ada... | ai_general | cs.AI | 2025-12-14T13:38:06+00:00 |
2512.23437v2 | RealX3D: A Physically-Degraded 3D Benchmark for Multi-view Visual Restoration and Reconstruction | We introduce RealX3D, a real-capture benchmark for multi-view visual restoration and 3D reconstruction under diverse physical degradations. RealX3D groups corruptions into four families, including illumination, scattering, occlusion, and blurring, and captures each at multiple severity levels using a unified acquisitio... | RealX3D: A Physically-Degraded 3D Benchmark for Multi-view Visual Restoration and Reconstruction. We introduce RealX3D, a real-capture benchmark for multi-view visual restoration and 3D reconstruction under diverse physical degradations. RealX3D groups corruptions into four families, including illumination, scattering,... | cv_vision | cs.CV | 2025-12-29T12:57:19+00:00 |
2512.24294v1 | Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction | Robust preprocessing is rarely quantified in deep-learning pipelines for low-dose CT (LDCT) lung cancer screening. We develop and validate Virtual-Eyes, a clinically motivated 16-bit CT quality-control pipeline, and measure its differential impact on generalist foundation models versus specialist models. Virtual-Eyes e... | Virtual-Eyes: Quantitative Validation of a Lung CT Quality-Control Pipeline for Foundation-Model Cancer Risk Prediction. Robust preprocessing is rarely quantified in deep-learning pipelines for low-dose CT (LDCT) lung cancer screening. We develop and validate Virtual-Eyes, a clinically motivated 16-bit CT quality-contr... | cv_vision | cs.CV | 2025-12-30T15:34:18+00:00 |
2510.05380v1 | Minima and Critical Points of the Bethe Free Energy Are Invariant Under Deformation Retractions of Factor Graphs | In graphical models, factor graphs, and more generally energy-based models, the interactions between variables are encoded by a graph, a hypergraph, or, in the most general case, a partially ordered set (poset). Inference on such probabilistic models cannot be performed exactly due to cycles in the underlying structure... | Minima and Critical Points of the Bethe Free Energy Are Invariant Under Deformation Retractions of Factor Graphs. In graphical models, factor graphs, and more generally energy-based models, the interactions between variables are encoded by a graph, a hypergraph, or, in the most general case, a partially ordered set (po... | stat_ml | stat.ML | 2025-10-06T21:16:31+00:00 |
2510.07832v1 | Surrogate Graph Partitioning for Spatial Prediction | Spatial prediction refers to the estimation of unobserved values from spatially distributed observations. Although recent advances have improved the capacity to model diverse observation types, adoption in practice remains limited in industries that demand interpretability. To mitigate this gap, surrogate models that e... | Surrogate Graph Partitioning for Spatial Prediction. Spatial prediction refers to the estimation of unobserved values from spatially distributed observations. Although recent advances have improved the capacity to model diverse observation types, adoption in practice remains limited in industries that demand interpreta... | stat_ml | stat.ML | 2025-10-09T06:24:49+00:00 |
2512.17566v1 | A unified FLAIR hyperintensity segmentation model for various CNS tumor types and acquisition time points | T2-weighted fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) scans are important for diagnosis, treatment planning and monitoring of brain tumors. Depending on the brain tumor type, the FLAIR hyperintensity volume is an important measure to asses the tumor volume or surrounding edema, and an... | A unified FLAIR hyperintensity segmentation model for various CNS tumor types and acquisition time points. T2-weighted fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) scans are important for diagnosis, treatment planning and monitoring of brain tumors. Depending on the brain tumor type, the... | cv_vision | cs.CV | 2025-12-19T13:33:43+00:00 |
2512.23426v1 | Direct Diffusion Score Preference Optimization via Stepwise Contrastive Policy-Pair Supervision | Diffusion models have achieved impressive results in generative tasks such as text-to-image synthesis, yet they often struggle to fully align outputs with nuanced user intent and maintain consistent aesthetic quality. Existing preference-based training methods like Diffusion Direct Preference Optimization help address ... | Direct Diffusion Score Preference Optimization via Stepwise Contrastive Policy-Pair Supervision. Diffusion models have achieved impressive results in generative tasks such as text-to-image synthesis, yet they often struggle to fully align outputs with nuanced user intent and maintain consistent aesthetic quality. Exist... | cv_vision | cs.CV | 2025-12-29T12:46:07+00:00 |
2510.15363v1 | Kernel Regression in Structured Non-IID Settings: Theory and Implications for Denoising Score Learning | Kernel ridge regression (KRR) is a foundational tool in machine learning, with recent work emphasizing its connections to neural networks. However, existing theory primarily addresses the i.i.d. setting, while real-world data often exhibits structured dependencies - particularly in applications like denoising score lea... | Kernel Regression in Structured Non-IID Settings: Theory and Implications for Denoising Score Learning. Kernel ridge regression (KRR) is a foundational tool in machine learning, with recent work emphasizing its connections to neural networks. However, existing theory primarily addresses the i.i.d. setting, while real-w... | stat_ml | stat.ML | 2025-10-17T06:52:40+00:00 |
2512.18177v1 | NEURO-GUARD: Neuro-Symbolic Generalization and Unbiased Adaptive Routing for Diagnostics -- Explainable Medical AI | Accurate yet interpretable image-based diagnosis remains a central challenge in medical AI, particularly in settings characterized by limited data, subtle visual cues, and high-stakes clinical decision-making. Most existing vision models rely on purely data-driven learning and produce black-box predictions with limited... | NEURO-GUARD: Neuro-Symbolic Generalization and Unbiased Adaptive Routing for Diagnostics -- Explainable Medical AI. Accurate yet interpretable image-based diagnosis remains a central challenge in medical AI, particularly in settings characterized by limited data, subtle visual cues, and high-stakes clinical decision-ma... | ai_general | cs.AI | 2025-12-20T02:32:15+00:00 |
2512.15146v4 | Beyond Majority Voting: Towards Fine-grained and More Reliable Reward Signal for Test-Time Reinforcement Learning | Test-time reinforcement learning mitigates the reliance on annotated data by using majority voting results as pseudo-labels, emerging as a complementary direction to reinforcement learning with verifiable rewards (RLVR) for improving reasoning ability. However, this voting strategy often induces confirmation bias and s... | Beyond Majority Voting: Towards Fine-grained and More Reliable Reward Signal for Test-Time Reinforcement Learning. Test-time reinforcement learning mitigates the reliance on annotated data by using majority voting results as pseudo-labels, emerging as a complementary direction to reinforcement learning with verifiable ... | cl_nlp | cs.CL | 2025-12-17T07:21:54+00:00 |
2512.14856v2 | T5Gemma 2: Seeing, Reading, and Understanding Longer | We introduce T5Gemma 2, the next generation of the T5Gemma family of lightweight open encoder-decoder models, featuring strong multilingual, multimodal and long-context capabilities. T5Gemma 2 follows the adaptation recipe (via UL2) in T5Gemma -- adapting a pretrained decoder-only model into an encoder-decoder model, a... | T5Gemma 2: Seeing, Reading, and Understanding Longer. We introduce T5Gemma 2, the next generation of the T5Gemma family of lightweight open encoder-decoder models, featuring strong multilingual, multimodal and long-context capabilities. T5Gemma 2 follows the adaptation recipe (via UL2) in T5Gemma -- adapting a pretrain... | cl_nlp | cs.CL | 2025-12-16T19:19:34+00:00 |
2512.22939v3 | ColaVLA: Leveraging Cognitive Latent Reasoning for Hierarchical Parallel Trajectory Planning in Autonomous Driving | Autonomous driving requires generating safe and reliable trajectories from complex multimodal inputs. Traditional modular pipelines separate perception, prediction, and planning, while recent end-to-end (E2E) systems learn them jointly. Vision-language models (VLMs) further enrich this paradigm by introducing cross-mod... | ColaVLA: Leveraging Cognitive Latent Reasoning for Hierarchical Parallel Trajectory Planning in Autonomous Driving. Autonomous driving requires generating safe and reliable trajectories from complex multimodal inputs. Traditional modular pipelines separate perception, prediction, and planning, while recent end-to-end (... | cv_vision | cs.CV | 2025-12-28T14:06:37+00:00 |
2512.20136v3 | M$^3$KG-RAG: Multi-hop Multimodal Knowledge Graph-enhanced Retrieval-Augmented Generation | Retrieval-Augmented Generation (RAG) has recently been extended to multimodal settings, connecting multimodal large language models (MLLMs) with vast corpora of external knowledge such as multimodal knowledge graphs (MMKGs). Despite their recent success, multimodal RAG in the audio-visual domain remains challenging due... | M$^3$KG-RAG: Multi-hop Multimodal Knowledge Graph-enhanced Retrieval-Augmented Generation. Retrieval-Augmented Generation (RAG) has recently been extended to multimodal settings, connecting multimodal large language models (MLLMs) with vast corpora of external knowledge such as multimodal knowledge graphs (MMKGs). Desp... | cl_nlp | cs.CL | 2025-12-23T07:54:03+00:00 |
2601.00864v1 | Distribution Matching for Graph Quantification Under Structural Covariate Shift | Graphs are commonly used in machine learning to model relationships between instances. Consider the task of predicting the political preferences of users in a social network; to solve this task one should consider, both, the features of each individual user and the relationships between them. However, oftentimes one is... | Distribution Matching for Graph Quantification Under Structural Covariate Shift. Graphs are commonly used in machine learning to model relationships between instances. Consider the task of predicting the political preferences of users in a social network; to solve this task one should consider, both, the features of ea... | lg_ml | cs.LG | 2025-12-30T11:58:33+00:00 |
2512.16762v3 | NRGPT: An Energy-based Alternative for GPT | Generative Pre-trained Transformer (GPT) architectures are the most popular design for language modeling. Energy-based modeling is a different paradigm that views inference as a dynamical process operating on an energy landscape. We propose a minimal modification of the GPT setting to unify it with the EBM framework. T... | NRGPT: An Energy-based Alternative for GPT. Generative Pre-trained Transformer (GPT) architectures are the most popular design for language modeling. Energy-based modeling is a different paradigm that views inference as a dynamical process operating on an energy landscape. We propose a minimal modification of the GPT s... | lg_ml | cs.LG | 2025-12-18T16:59:10+00:00 |
2512.22199v1 | Bidirectional RAG: Safe Self-Improving Retrieval-Augmented Generation Through Multi-Stage Validation | Retrieval-Augmented Generation RAG systems enhance large language models by grounding responses in external knowledge bases, but conventional RAG architectures operate with static corpora that cannot evolve from user interactions. We introduce Bidirectional RAG, a novel RAG architecture that enables safe corpus expansi... | Bidirectional RAG: Safe Self-Improving Retrieval-Augmented Generation Through Multi-Stage Validation. Retrieval-Augmented Generation RAG systems enhance large language models by grounding responses in external knowledge bases, but conventional RAG architectures operate with static corpora that cannot evolve from user i... | ai_general | cs.AI | 2025-12-20T19:42:42+00:00 |
2512.16144v1 | INTELLECT-3: Technical Report | We present INTELLECT-3, a 106B-parameter Mixture-of-Experts model (12B active) trained with large-scale reinforcement learning on our end-to-end RL infrastructure stack. INTELLECT-3 achieves state of the art performance for its size across math, code, science and reasoning benchmarks, outperforming many larger frontier... | INTELLECT-3: Technical Report. We present INTELLECT-3, a 106B-parameter Mixture-of-Experts model (12B active) trained with large-scale reinforcement learning on our end-to-end RL infrastructure stack. INTELLECT-3 achieves state of the art performance for its size across math, code, science and reasoning benchmarks, out... | lg_ml | cs.LG | 2025-12-18T03:57:01+00:00 |
2512.17532v1 | Robust-R1: Degradation-Aware Reasoning for Robust Visual Understanding | Multimodal Large Language Models struggle to maintain reliable performance under extreme real-world visual degradations, which impede their practical robustness. Existing robust MLLMs predominantly rely on implicit training/adaptation that focuses solely on visual encoder generalization, suffering from limited interpre... | Robust-R1: Degradation-Aware Reasoning for Robust Visual Understanding. Multimodal Large Language Models struggle to maintain reliable performance under extreme real-world visual degradations, which impede their practical robustness. Existing robust MLLMs predominantly rely on implicit training/adaptation that focuses ... | cv_vision | cs.CV | 2025-12-19T12:56:17+00:00 |
2509.17359v2 | MLLM-Driven Semantic Identifier Generation for Generative Cross-Modal Retrieval | Generative cross-modal retrieval, which treats retrieval as a generation task, has emerged as a promising direction with the rise of Multimodal Large Language Models (MLLMs). In this setting, the model responds to a text query by generating an identifier corresponding to the target image. However, existing methods typi... | MLLM-Driven Semantic Identifier Generation for Generative Cross-Modal Retrieval. Generative cross-modal retrieval, which treats retrieval as a generation task, has emerged as a promising direction with the rise of Multimodal Large Language Models (MLLMs). In this setting, the model responds to a text query by generatin... | ir_retrieval | cs.IR | 2025-09-22T05:23:06+00:00 |
2512.04367v1 | AgentBay: A Hybrid Interaction Sandbox for Seamless Human-AI Intervention in Agentic Systems | The rapid advancement of Large Language Models (LLMs) is catalyzing a shift towards autonomous AI Agents capable of executing complex, multi-step tasks. However, these agents remain brittle when faced with real-world exceptions, making Human-in-the-Loop (HITL) supervision essential for mission-critical applications. In... | AgentBay: A Hybrid Interaction Sandbox for Seamless Human-AI Intervention in Agentic Systems. The rapid advancement of Large Language Models (LLMs) is catalyzing a shift towards autonomous AI Agents capable of executing complex, multi-step tasks. However, these agents remain brittle when faced with real-world exception... | ai_general | cs.AI | 2025-12-04T01:31:00+00:00 |
2512.06653v3 | LightSearcher: Efficient DeepSearch via Experiential Memory | DeepSearch paradigms have become a core enabler for deep reasoning models, allowing them to invoke external search tools to access up-to-date, domain-specific knowledge beyond parametric boundaries, thereby enhancing the depth and factual reliability of reasoning. Building upon this foundation, recent advances in reinf... | LightSearcher: Efficient DeepSearch via Experiential Memory. DeepSearch paradigms have become a core enabler for deep reasoning models, allowing them to invoke external search tools to access up-to-date, domain-specific knowledge beyond parametric boundaries, thereby enhancing the depth and factual reliability of reaso... | ai_general | cs.AI | 2025-12-07T04:29:52+00:00 |
2512.11052v1 | An Efficient Variant of One-Class SVM with Lifelong Online Learning Guarantees | We study outlier (a.k.a., anomaly) detection for single-pass non-stationary streaming data. In the well-studied offline or batch outlier detection problem, traditional methods such as kernel One-Class SVM (OCSVM) are both computationally heavy and prone to large false-negative (Type II) errors under non-stationarity. T... | An Efficient Variant of One-Class SVM with Lifelong Online Learning Guarantees. We study outlier (a.k.a., anomaly) detection for single-pass non-stationary streaming data. In the well-studied offline or batch outlier detection problem, traditional methods such as kernel One-Class SVM (OCSVM) are both computationally he... | stat_ml | stat.ML | 2025-12-11T19:09:58+00:00 |
2509.02943v1 | Knowledge graph-based personalized multimodal recommendation fusion framework | In the contemporary age characterized by information abundance, rapid advancements in artificial intelligence have rendered recommendation systems indispensable. Conventional recommendation methodologies based on collaborative filtering or individual attributes encounter deficiencies in capturing nuanced user interests... | Knowledge graph-based personalized multimodal recommendation fusion framework. In the contemporary age characterized by information abundance, rapid advancements in artificial intelligence have rendered recommendation systems indispensable. Conventional recommendation methodologies based on collaborative filtering or i... | ir_retrieval | cs.IR | 2025-09-03T02:17:28+00:00 |
2512.25052v1 | AdaGReS:Adaptive Greedy Context Selection via Redundancy-Aware Scoring for Token-Budgeted RAG | Retrieval-augmented generation (RAG) is highly sensitive to the quality of selected context, yet standard top-k retrieval often returns redundant or near-duplicate chunks that waste token budget and degrade downstream generation. We present AdaGReS, a redundancy-aware context selection framework for token-budgeted RAG ... | AdaGReS:Adaptive Greedy Context Selection via Redundancy-Aware Scoring for Token-Budgeted RAG. Retrieval-augmented generation (RAG) is highly sensitive to the quality of selected context, yet standard top-k retrieval often returns redundant or near-duplicate chunks that waste token budget and degrade downstream generat... | cl_nlp | cs.CL | 2025-12-31T18:48:07+00:00 |
2512.21792v1 | AI for Mycetoma Diagnosis in Histopathological Images: The MICCAI 2024 Challenge | Mycetoma is a neglected tropical disease caused by fungi or bacteria leading to severe tissue damage and disabilities. It affects poor and rural communities and presents medical challenges and socioeconomic burdens on patients and healthcare systems in endemic regions worldwide. Mycetoma diagnosis is a major challenge ... | AI for Mycetoma Diagnosis in Histopathological Images: The MICCAI 2024 Challenge. Mycetoma is a neglected tropical disease caused by fungi or bacteria leading to severe tissue damage and disabilities. It affects poor and rural communities and presents medical challenges and socioeconomic burdens on patients and healthc... | cv_vision | cs.CV | 2025-12-25T21:46:33+00:00 |
2512.23245v2 | ASemConsist: Adaptive Semantic Feature Control for Training-Free Identity-Consistent Generation | Recent text-to-image diffusion models have significantly improved visual quality and text alignment. However, generating a sequence of images while preserving consistent character identity across diverse scene descriptions remains a challenging task. Existing methods often struggle with a trade-off between maintaining ... | ASemConsist: Adaptive Semantic Feature Control for Training-Free Identity-Consistent Generation. Recent text-to-image diffusion models have significantly improved visual quality and text alignment. However, generating a sequence of images while preserving consistent character identity across diverse scene descriptions ... | cv_vision | cs.CV | 2025-12-29T07:06:57+00:00 |
2512.19027v2 | Recontextualization Mitigates Specification Gaming without Modifying the Specification | Developers often struggle to specify correct training labels and rewards. Perhaps they don't need to. We propose recontextualization, which reduces how often language models "game" training signals, performing misbehaviors those signals mistakenly reinforce. We show recontextualization prevents models from learning to ... | Recontextualization Mitigates Specification Gaming without Modifying the Specification. Developers often struggle to specify correct training labels and rewards. Perhaps they don't need to. We propose recontextualization, which reduces how often language models "game" training signals, performing misbehaviors those sig... | ai_general | cs.AI | 2025-12-22T04:53:40+00:00 |
2512.19991v2 | Bloom Filter Encoding for Machine Learning | We present a method that uses a Bloom filter transform to preprocess data for machine learning. Each sample is encoded into a compact bit-array representation using hash-based encoding, producing a fixed-length feature space that reduces memory usage and obfuscates original feature values. The encoding does not rely on... | Bloom Filter Encoding for Machine Learning. We present a method that uses a Bloom filter transform to preprocess data for machine learning. Each sample is encoded into a compact bit-array representation using hash-based encoding, producing a fixed-length feature space that reduces memory usage and obfuscates original f... | lg_ml | cs.LG | 2025-12-23T02:33:57+00:00 |
2511.16106v1 | Incorporating Token Importance in Multi-Vector Retrieval | ColBERT introduced a late interaction mechanism that independently encodes queries and documents using BERT, and computes similarity via fine-grained interactions over token-level vector representations. This design enables expressive matching while allowing efficient computation of scores, as the multi-vector document... | Incorporating Token Importance in Multi-Vector Retrieval. ColBERT introduced a late interaction mechanism that independently encodes queries and documents using BERT, and computes similarity via fine-grained interactions over token-level vector representations. This design enables expressive matching while allowing eff... | ir_retrieval | cs.IR | 2025-11-20T06:58:31+00:00 |
2512.21315v1 | Does the Data Processing Inequality Reflect Practice? On the Utility of Low-Level Tasks | The data processing inequality is an information-theoretic principle stating that the information content of a signal cannot be increased by processing the observations. In particular, it suggests that there is no benefit in enhancing the signal or encoding it before addressing a classification problem. This assertion ... | Does the Data Processing Inequality Reflect Practice? On the Utility of Low-Level Tasks. The data processing inequality is an information-theoretic principle stating that the information content of a signal cannot be increased by processing the observations. In particular, it suggests that there is no benefit in enhanc... | lg_ml | cs.LG | 2025-12-24T18:21:01+00:00 |
2510.14074v1 | Exact Dynamics of Multi-class Stochastic Gradient Descent | We develop a framework for analyzing the training and learning rate dynamics on a variety of high- dimensional optimization problems trained using one-pass stochastic gradient descent (SGD) with data generated from multiple anisotropic classes. We give exact expressions for a large class of functions of the limiting dy... | Exact Dynamics of Multi-class Stochastic Gradient Descent. We develop a framework for analyzing the training and learning rate dynamics on a variety of high- dimensional optimization problems trained using one-pass stochastic gradient descent (SGD) with data generated from multiple anisotropic classes. We give exact ex... | stat_ml | stat.ML | 2025-10-15T20:31:49+00:00 |
2512.09148v1 | Detecting Hallucinations in Graph Retrieval-Augmented Generation via Attention Patterns and Semantic Alignment | Graph-based Retrieval-Augmented Generation (GraphRAG) enhances Large Language Models (LLMs) by incorporating external knowledge from linearized subgraphs retrieved from knowledge graphs. However, LLMs struggle to interpret the relational and topological information in these inputs, resulting in hallucinations that are ... | Detecting Hallucinations in Graph Retrieval-Augmented Generation via Attention Patterns and Semantic Alignment. Graph-based Retrieval-Augmented Generation (GraphRAG) enhances Large Language Models (LLMs) by incorporating external knowledge from linearized subgraphs retrieved from knowledge graphs. However, LLMs struggl... | cl_nlp | cs.CL | 2025-12-09T21:52:50+00:00 |
2509.18484v2 | Estimating Heterogeneous Causal Effect on Networks via Orthogonal Learning | Estimating causal effects on networks is challenging because treatments may affect both treated units and their neighbors, while network homophily induces dependence and confounding. These challenges are amplified when causal effects are heterogeneous across units and edges. We propose a two-stage orthogonal learning f... | Estimating Heterogeneous Causal Effect on Networks via Orthogonal Learning. Estimating causal effects on networks is challenging because treatments may affect both treated units and their neighbors, while network homophily induces dependence and confounding. These challenges are amplified when causal effects are hetero... | stat_ml | stat.ML | 2025-09-23T00:41:04+00:00 |
2512.12654v1 | Modeling Authorial Style in Urdu Novels Using Character Interaction Graphs and Graph Neural Networks | Authorship analysis has traditionally focused on lexical and stylistic cues within text, while higher-level narrative structure remains underexplored, particularly for low-resource languages such as Urdu. This work proposes a graph-based framework that models Urdu novels as character interaction networks to examine whe... | Modeling Authorial Style in Urdu Novels Using Character Interaction Graphs and Graph Neural Networks. Authorship analysis has traditionally focused on lexical and stylistic cues within text, while higher-level narrative structure remains underexplored, particularly for low-resource languages such as Urdu. This work pro... | cl_nlp | cs.CL | 2025-12-14T11:59:16+00:00 |
2512.22612v1 | Enhancing Noise Resilience in Face Clustering via Sparse Differential Transformer | The method used to measure relationships between face embeddings plays a crucial role in determining the performance of face clustering. Existing methods employ the Jaccard similarity coefficient instead of the cosine distance to enhance the measurement accuracy. However, these methods introduce too many irrelevant nod... | Enhancing Noise Resilience in Face Clustering via Sparse Differential Transformer. The method used to measure relationships between face embeddings plays a crucial role in determining the performance of face clustering. Existing methods employ the Jaccard similarity coefficient instead of the cosine distance to enhance... | cv_vision | cs.CV | 2025-12-27T14:39:43+00:00 |
2512.14620v1 | JMMMU-Pro: Image-based Japanese Multi-discipline Multimodal Understanding Benchmark via Vibe Benchmark Construction | This paper introduces JMMMU-Pro, an image-based Japanese Multi-discipline Multimodal Understanding Benchmark, and Vibe Benchmark Construction, a scalable construction method. Following the evolution from MMMU to MMMU-Pro, JMMMU-Pro extends JMMMU by composing the question image and question text into a single image, the... | JMMMU-Pro: Image-based Japanese Multi-discipline Multimodal Understanding Benchmark via Vibe Benchmark Construction. This paper introduces JMMMU-Pro, an image-based Japanese Multi-discipline Multimodal Understanding Benchmark, and Vibe Benchmark Construction, a scalable construction method. Following the evolution from... | cl_nlp | cs.CL | 2025-12-16T17:33:00+00:00 |
2509.07269v1 | Datasets for Navigating Sensitive Topics in Recommendation Systems | Personalized AI systems, from recommendation systems to chatbots, are a prevalent method for distributing content to users based on their learned preferences. However, there is growing concern about the adverse effects of these systems, including their potential tendency to expose users to sensitive or harmful material... | Datasets for Navigating Sensitive Topics in Recommendation Systems. Personalized AI systems, from recommendation systems to chatbots, are a prevalent method for distributing content to users based on their learned preferences. However, there is growing concern about the adverse effects of these systems, including their... | ir_retrieval | cs.IR | 2025-09-08T22:58:17+00:00 |
2601.11559v1 | MIMIC-RD: Can LLMs differentially diagnose rare diseases in real-world clinical settings? | Despite rare diseases affecting 1 in 10 Americans, their differential diagnosis remains challenging. Due to their impressive recall abilities, large language models (LLMs) have been recently explored for differential diagnosis. Existing approaches to evaluating LLM-based rare disease diagnosis suffer from two critical ... | MIMIC-RD: Can LLMs differentially diagnose rare diseases in real-world clinical settings?. Despite rare diseases affecting 1 in 10 Americans, their differential diagnosis remains challenging. Due to their impressive recall abilities, large language models (LLMs) have been recently explored for differential diagnosis. E... | ai_general | cs.AI | 2025-12-18T05:03:49+00:00 |
2512.18614v1 | PTTA: A Pure Text-to-Animation Framework for High-Quality Creation | Traditional animation production involves complex pipelines and significant manual labor cost. While recent video generation models such as Sora, Kling, and CogVideoX achieve impressive results on natural video synthesis, they exhibit notable limitations when applied to animation generation. Recent efforts, such as Ani... | PTTA: A Pure Text-to-Animation Framework for High-Quality Creation. Traditional animation production involves complex pipelines and significant manual labor cost. While recent video generation models such as Sora, Kling, and CogVideoX achieve impressive results on natural video synthesis, they exhibit notable limitatio... | cv_vision | cs.CV | 2025-12-21T06:17:28+00:00 |
2512.04492v1 | MSME: A Multi-Stage Multi-Expert Framework for Zero-Shot Stance Detection | LLM-based approaches have recently achieved impressive results in zero-shot stance detection. However, they still struggle in complex real-world scenarios, where stance understanding requires dynamic background knowledge, target definitions involve compound entities or events that must be explicitly linked to stance la... | MSME: A Multi-Stage Multi-Expert Framework for Zero-Shot Stance Detection. LLM-based approaches have recently achieved impressive results in zero-shot stance detection. However, they still struggle in complex real-world scenarios, where stance understanding requires dynamic background knowledge, target definitions invo... | cl_nlp | cs.CL | 2025-12-04T05:56:13+00:00 |
2512.23421v3 | DriveLaW:Unifying Planning and Video Generation in a Latent Driving World | World models have become crucial for autonomous driving, as they learn how scenarios evolve over time to address the long-tail challenges of the real world. However, current approaches relegate world models to limited roles: they operate within ostensibly unified architectures that still keep world prediction and motio... | DriveLaW:Unifying Planning and Video Generation in a Latent Driving World. World models have become crucial for autonomous driving, as they learn how scenarios evolve over time to address the long-tail challenges of the real world. However, current approaches relegate world models to limited roles: they operate within ... | cv_vision | cs.CV | 2025-12-29T12:32:27+00:00 |
2512.24259v1 | Tracing the Flow of Knowledge From Science to Technology Using Deep Learning | We develop a language similarity model suitable for working with patents and scientific publications at the same time. In a horse race-style evaluation, we subject eight language (similarity) models to predict credible Patent-Paper Citations. We find that our Pat-SPECTER model performs best, which is the SPECTER2 model... | Tracing the Flow of Knowledge From Science to Technology Using Deep Learning. We develop a language similarity model suitable for working with patents and scientific publications at the same time. In a horse race-style evaluation, we subject eight language (similarity) models to predict credible Patent-Paper Citations.... | cl_nlp | cs.CL | 2025-12-30T14:36:17+00:00 |
2511.02373v1 | A new class of Markov random fields enabling lightweight sampling | This work addresses the problem of efficient sampling of Markov random fields (MRF). The sampling of Potts or Ising MRF is most often based on Gibbs sampling, and is thus computationally expensive. We consider in this work how to circumvent this bottleneck through a link with Gaussian Markov Random fields. The latter c... | A new class of Markov random fields enabling lightweight sampling. This work addresses the problem of efficient sampling of Markov random fields (MRF). The sampling of Potts or Ising MRF is most often based on Gibbs sampling, and is thus computationally expensive. We consider in this work how to circumvent this bottlen... | stat_ml | stat.ML | 2025-11-04T08:53:17+00:00 |
2512.06732v1 | "The Dentist is an involved parent, the bartender is not": Revealing Implicit Biases in QA with Implicit BBQ | Existing benchmarks evaluating biases in large language models (LLMs) primarily rely on explicit cues, declaring protected attributes like religion, race, gender by name. However, real-world interactions often contain implicit biases, inferred subtly through names, cultural cues, or traits. This critical oversight crea... | "The Dentist is an involved parent, the bartender is not": Revealing Implicit Biases in QA with Implicit BBQ. Existing benchmarks evaluating biases in large language models (LLMs) primarily rely on explicit cues, declaring protected attributes like religion, race, gender by name. However, real-world interactions often ... | cl_nlp | cs.CL | 2025-12-07T08:57:27+00:00 |
2511.06405v2 | TOOL4POI: A Tool-Augmented LLM Framework for Next POI Recommendation | Next Point-of-Interest (POI) recommendation is a fundamental task in location-based services. While recent advances leverage Large Language Model (LLM) for sequential modeling, existing LLM-based approaches face two key limitations: (i) strong reliance on the contextual completeness of user histories, resulting in poor... | TOOL4POI: A Tool-Augmented LLM Framework for Next POI Recommendation. Next Point-of-Interest (POI) recommendation is a fundamental task in location-based services. While recent advances leverage Large Language Model (LLM) for sequential modeling, existing LLM-based approaches face two key limitations: (i) strong relian... | ir_retrieval | cs.IR | 2025-11-09T14:37:11+00:00 |
2512.24975v1 | Attribution-Guided Distillation of Matryoshka Sparse Autoencoders | Sparse autoencoders (SAEs) aim to disentangle model activations into monosemantic, human-interpretable features. In practice, learned features are often redundant and vary across training runs and sparsity levels, which makes interpretations difficult to transfer and reuse. We introduce Distilled Matryoshka Sparse Auto... | Attribution-Guided Distillation of Matryoshka Sparse Autoencoders. Sparse autoencoders (SAEs) aim to disentangle model activations into monosemantic, human-interpretable features. In practice, learned features are often redundant and vary across training runs and sparsity levels, which makes interpretations difficult t... | lg_ml | cs.LG | 2025-12-31T17:12:55+00:00 |
2601.05268v1 | Separating Semantic Expansion from Linear Geometry for PubMed-Scale Vector Search | We describe a PubMed scale retrieval framework that separates semantic interpretation from metric geometry. A large language model expands a natural language query into concise biomedical phrases; retrieval then operates in a fixed, mean free, approximately isotropic embedding space. Each document and query vector is f... | Separating Semantic Expansion from Linear Geometry for PubMed-Scale Vector Search. We describe a PubMed scale retrieval framework that separates semantic interpretation from metric geometry. A large language model expands a natural language query into concise biomedical phrases; retrieval then operates in a fixed, mean... | ir_retrieval | cs.IR | 2025-11-14T17:55:16+00:00 |
2512.24613v1 | Group Deliberation Oriented Multi-Agent Conversational Model for Complex Reasoning | This paper proposes a group deliberation oriented multi-agent conversational model to address the limitations of single large language models in complex reasoning tasks. The model adopts a three-level role division architecture consisting of generation, verification, and integration. An opinion generation agent produce... | Group Deliberation Oriented Multi-Agent Conversational Model for Complex Reasoning. This paper proposes a group deliberation oriented multi-agent conversational model to address the limitations of single large language models in complex reasoning tasks. The model adopts a three-level role division architecture consisti... | ai_general | cs.AI | 2025-12-31T04:10:57+00:00 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.