id
string
title
string
abstract
string
text
string
label
string
primary_category
string
published
string
2512.24338v2
The Mechanics of CNN Filtering with Rectification
This paper proposes elementary information mechanics as a new model for understanding the mechanical properties of convolutional filtering with rectification, inspired by physical theories of special relativity and quantum mechanics. We consider kernels decomposed into orthogonal even and odd components. Even component...
The Mechanics of CNN Filtering with Rectification. This paper proposes elementary information mechanics as a new model for understanding the mechanical properties of convolutional filtering with rectification, inspired by physical theories of special relativity and quantum mechanics. We consider kernels decomposed into...
cv_vision
cs.CV
2025-12-30T16:44:41+00:00
2512.09727v1
Gaussian Process Aggregation for Root-Parallel Monte Carlo Tree Search with Continuous Actions
Monte Carlo Tree Search is a cornerstone algorithm for online planning, and its root-parallel variant is widely used when wall clock time is limited but best performance is desired. In environments with continuous action spaces, how to best aggregate statistics from different threads is an important yet underexplored q...
Gaussian Process Aggregation for Root-Parallel Monte Carlo Tree Search with Continuous Actions. Monte Carlo Tree Search is a cornerstone algorithm for online planning, and its root-parallel variant is widely used when wall clock time is limited but best performance is desired. In environments with continuous action spa...
ai_general
cs.AI
2025-12-10T15:09:06+00:00
2512.14792v1
IaC Generation with LLMs: An Error Taxonomy and A Study on Configuration Knowledge Injection
Large Language Models (LLMs) currently exhibit low success rates in generating correct and intent-aligned Infrastructure as Code (IaC). This research investigated methods to improve LLM-based IaC generation, specifically for Terraform, by systematically injecting structured configuration knowledge. To facilitate this, ...
IaC Generation with LLMs: An Error Taxonomy and A Study on Configuration Knowledge Injection. Large Language Models (LLMs) currently exhibit low success rates in generating correct and intent-aligned Infrastructure as Code (IaC). This research investigated methods to improve LLM-based IaC generation, specifically for T...
ai_general
cs.AI
2025-12-16T14:58:00+00:00
2512.14228v1
Georeferencing complex relative locality descriptions with large language models
Georeferencing text documents has typically relied on either gazetteer-based methods to assign geographic coordinates to place names, or on language modelling approaches that associate textual terms with geographic locations. However, many location descriptions specify positions relatively with spatial relationships, m...
Georeferencing complex relative locality descriptions with large language models. Georeferencing text documents has typically relied on either gazetteer-based methods to assign geographic coordinates to place names, or on language modelling approaches that associate textual terms with geographic locations. However, man...
ai_general
cs.AI
2025-12-16T09:27:02+00:00
2512.07538v3
SwissGov-RSD: A Human-annotated, Cross-lingual Benchmark for Token-level Recognition of Semantic Differences Between Related Documents
Recognizing semantic differences across documents is crucial for text generation evaluation and content alignment, especially in cross-lingual settings. However, as a standalone task, it has received little attention. We address this by introducing SwissGov-RSD, the first naturalistic, document-level, cross-lingual dat...
SwissGov-RSD: A Human-annotated, Cross-lingual Benchmark for Token-level Recognition of Semantic Differences Between Related Documents. Recognizing semantic differences across documents is crucial for text generation evaluation and content alignment, especially in cross-lingual settings. However, as a standalone task, ...
cl_nlp
cs.CL
2025-12-08T13:17:27+00:00
2512.19609v1
MapTrace: Scalable Data Generation for Route Tracing on Maps
While Multimodal Large Language Models have achieved human-like performance on many visual and textual reasoning tasks, their proficiency in fine-grained spatial understanding, such as route tracing on maps remains limited. Unlike humans, who can quickly learn to parse and navigate maps, current models often fail to re...
MapTrace: Scalable Data Generation for Route Tracing on Maps. While Multimodal Large Language Models have achieved human-like performance on many visual and textual reasoning tasks, their proficiency in fine-grained spatial understanding, such as route tracing on maps remains limited. Unlike humans, who can quickly lea...
cv_vision
cs.CV
2025-12-22T17:45:39+00:00
2512.15824v2
State-Augmented Graphs for Circular Economy Triage
Circular economy (CE) triage is the assessment of products to determine which sustainable pathway they can follow once they reach the end of their usefulness as they are currently being used. Effective CE triage requires adaptive decisions that balance retained value against the costs and constraints of processing and ...
State-Augmented Graphs for Circular Economy Triage. Circular economy (CE) triage is the assessment of products to determine which sustainable pathway they can follow once they reach the end of their usefulness as they are currently being used. Effective CE triage requires adaptive decisions that balance retained value ...
ai_general
cs.AI
2025-12-17T16:23:47+00:00
2512.18741v2
Memorize-and-Generate: Towards Long-Term Consistency in Real-Time Video Generation
Frame-level autoregressive (frame-AR) models have achieved significant progress, enabling real-time video generation comparable to bidirectional diffusion models and serving as a foundation for interactive world models and game engines. However, current approaches in long video generation typically rely on window atten...
Memorize-and-Generate: Towards Long-Term Consistency in Real-Time Video Generation. Frame-level autoregressive (frame-AR) models have achieved significant progress, enabling real-time video generation comparable to bidirectional diffusion models and serving as a foundation for interactive world models and game engines....
cv_vision
cs.CV
2025-12-21T14:02:53+00:00
2512.14491v1
Sparse Multi-Modal Transformer with Masking for Alzheimer's Disease Classification
Transformer-based multi-modal intelligent systems often suffer from high computational and energy costs due to dense self-attention, limiting their scalability under resource constraints. This paper presents SMMT, a sparse multi-modal transformer architecture designed to improve efficiency and robustness. Building upon...
Sparse Multi-Modal Transformer with Masking for Alzheimer's Disease Classification. Transformer-based multi-modal intelligent systems often suffer from high computational and energy costs due to dense self-attention, limiting their scalability under resource constraints. This paper presents SMMT, a sparse multi-modal t...
ai_general
cs.AI
2025-12-16T15:24:57+00:00
2512.16922v2
Next-Embedding Prediction Makes Strong Vision Learners
Inspired by the success of generative pretraining in natural language, we ask whether the same principles can yield strong self-supervised visual learners. Instead of training models to output features for downstream use, we train them to generate embeddings to perform predictive tasks directly. This work explores such...
Next-Embedding Prediction Makes Strong Vision Learners. Inspired by the success of generative pretraining in natural language, we ask whether the same principles can yield strong self-supervised visual learners. Instead of training models to output features for downstream use, we train them to generate embeddings to pe...
cv_vision
cs.CV
2025-12-18T18:59:58+00:00
2512.21694v1
BeHGAN: Bengali Handwritten Word Generation from Plain Text Using Generative Adversarial Networks
Handwritten Text Recognition (HTR) is a well-established research area. In contrast, Handwritten Text Generation (HTG) is an emerging field with significant potential. This task is challenging due to the variation in individual handwriting styles. A large and diverse dataset is required to generate realistic handwritte...
BeHGAN: Bengali Handwritten Word Generation from Plain Text Using Generative Adversarial Networks. Handwritten Text Recognition (HTR) is a well-established research area. In contrast, Handwritten Text Generation (HTG) is an emerging field with significant potential. This task is challenging due to the variation in indi...
cv_vision
cs.CV
2025-12-25T14:38:12+00:00
2512.13323v1
Error-Driven Prompt Optimization for Arithmetic Reasoning
Recent advancements in artificial intelligence have sparked interest in industrial agents capable of supporting analysts in regulated sectors, such as finance and healthcare, within tabular data workflows. A key capability for such systems is performing accurate arithmetic operations on structured data while ensuring s...
Error-Driven Prompt Optimization for Arithmetic Reasoning. Recent advancements in artificial intelligence have sparked interest in industrial agents capable of supporting analysts in regulated sectors, such as finance and healthcare, within tabular data workflows. A key capability for such systems is performing accurat...
ai_general
cs.AI
2025-12-15T13:39:14+00:00
2512.18046v1
YolovN-CBi: A Lightweight and Efficient Architecture for Real-Time Detection of Small UAVs
Unmanned Aerial Vehicles, commonly known as, drones pose increasing risks in civilian and defense settings, demanding accurate and real-time drone detection systems. However, detecting drones is challenging because of their small size, rapid movement, and low visual contrast. A modified architecture of YolovN called th...
YolovN-CBi: A Lightweight and Efficient Architecture for Real-Time Detection of Small UAVs. Unmanned Aerial Vehicles, commonly known as, drones pose increasing risks in civilian and defense settings, demanding accurate and real-time drone detection systems. However, detecting drones is challenging because of their smal...
cv_vision
cs.CV
2025-12-19T20:27:53+00:00
2509.20225v1
Multimodal Representation-disentangled Information Bottleneck for Multimodal Recommendation
Multimodal data has significantly advanced recommendation systems by integrating diverse information sources to model user preferences and item characteristics. However, these systems often struggle with redundant and irrelevant information, which can degrade performance. Most existing methods either fuse multimodal in...
Multimodal Representation-disentangled Information Bottleneck for Multimodal Recommendation. Multimodal data has significantly advanced recommendation systems by integrating diverse information sources to model user preferences and item characteristics. However, these systems often struggle with redundant and irrelevan...
ir_retrieval
cs.IR
2025-09-24T15:18:32+00:00
2509.19988v2
BioBO: Biology-informed Bayesian Optimization for Perturbation Design
Efficient design of genomic perturbation experiments is crucial for accelerating drug discovery and therapeutic target identification, yet exhaustive perturbation of the human genome remains infeasible due to the vast search space of potential genetic interactions and experimental constraints. Bayesian optimization (BO...
BioBO: Biology-informed Bayesian Optimization for Perturbation Design. Efficient design of genomic perturbation experiments is crucial for accelerating drug discovery and therapeutic target identification, yet exhaustive perturbation of the human genome remains infeasible due to the vast search space of potential genet...
stat_ml
stat.ML
2025-09-24T10:50:06+00:00
2512.23089v1
MedSAM-based lung masking for multi-label chest X-ray classification
Chest X-ray (CXR) imaging is widely used for screening and diagnosing pulmonary abnormalities, yet automated interpretation remains challenging due to weak disease signals, dataset bias, and limited spatial supervision. Foundation models for medical image segmentation (MedSAM) provide an opportunity to introduce anatom...
MedSAM-based lung masking for multi-label chest X-ray classification. Chest X-ray (CXR) imaging is widely used for screening and diagnosing pulmonary abnormalities, yet automated interpretation remains challenging due to weak disease signals, dataset bias, and limited spatial supervision. Foundation models for medical ...
cv_vision
cs.CV
2025-12-28T21:56:41+00:00
2512.09444v1
Advancing Text Classification with Large Language Models and Neural Attention Mechanisms
This study proposes a text classification algorithm based on large language models, aiming to address the limitations of traditional methods in capturing long-range dependencies, understanding contextual semantics, and handling class imbalance. The framework includes text encoding, contextual representation modeling, a...
Advancing Text Classification with Large Language Models and Neural Attention Mechanisms. This study proposes a text classification algorithm based on large language models, aiming to address the limitations of traditional methods in capturing long-range dependencies, understanding contextual semantics, and handling cl...
cl_nlp
cs.CL
2025-12-10T09:18:41+00:00
2510.00908v1
Bridging Language Gaps: Advances in Cross-Lingual Information Retrieval with Multilingual LLMs
Cross-lingual information retrieval (CLIR) addresses the challenge of retrieving relevant documents written in languages different from that of the original query. Research in this area has typically framed the task as monolingual retrieval augmented by translation, treating retrieval methods and cross-lingual capabili...
Bridging Language Gaps: Advances in Cross-Lingual Information Retrieval with Multilingual LLMs. Cross-lingual information retrieval (CLIR) addresses the challenge of retrieving relevant documents written in languages different from that of the original query. Research in this area has typically framed the task as monol...
ir_retrieval
cs.IR
2025-10-01T13:50:05+00:00
2512.14527v1
Dynamic Learning Rate Scheduling based on Loss Changes Leads to Faster Convergence
Despite significant advances in optimizers for training, most research works use common scheduler choices like Cosine or exponential decay. In this paper, we study \emph{GreedyLR}, a novel scheduler that adaptively adjusts the learning rate during training based on the current loss. To validate the effectiveness of our...
Dynamic Learning Rate Scheduling based on Loss Changes Leads to Faster Convergence. Despite significant advances in optimizers for training, most research works use common scheduler choices like Cosine or exponential decay. In this paper, we study \emph{GreedyLR}, a novel scheduler that adaptively adjusts the learning ...
ai_general
cs.AI
2025-12-16T16:03:52+00:00
2509.06472v2
Rethinking LLM Parametric Knowledge as Post-retrieval Confidence for Dynamic Retrieval and Reranking
Large Language Models (LLMs) often generate inaccurate responses (hallucinations) when faced with questions beyond their knowledge scope. Retrieval-Augmented Generation (RAG) addresses this by leveraging external knowledge, but a critical challenge remains: determining whether retrieved contexts effectively enhance the...
Rethinking LLM Parametric Knowledge as Post-retrieval Confidence for Dynamic Retrieval and Reranking. Large Language Models (LLMs) often generate inaccurate responses (hallucinations) when faced with questions beyond their knowledge scope. Retrieval-Augmented Generation (RAG) addresses this by leveraging external knowl...
ir_retrieval
cs.IR
2025-09-08T09:37:20+00:00
2512.05930v1
PRiSM: An Agentic Multimodal Benchmark for Scientific Reasoning via Python-Grounded Evaluation
Evaluating vision-language models (VLMs) in scientific domains like mathematics and physics poses unique challenges that go far beyond predicting final answers. These domains demand conceptual understanding, symbolic reasoning, and adherence to formal laws, requirements that most existing benchmarks fail to address. In...
PRiSM: An Agentic Multimodal Benchmark for Scientific Reasoning via Python-Grounded Evaluation. Evaluating vision-language models (VLMs) in scientific domains like mathematics and physics poses unique challenges that go far beyond predicting final answers. These domains demand conceptual understanding, symbolic reasoni...
ai_general
cs.AI
2025-12-05T18:14:55+00:00
2512.07543v2
Over-representation of phonological features in basic vocabulary doesn't replicate when controlling for spatial and phylogenetic effects
The statistical over-representation of phonological features in the basic vocabulary of languages is often interpreted as reflecting potentially universal sound symbolic patterns. However, most of those results have not been tested explicitly for reproducibility and might be prone to biases in the study samples or mode...
Over-representation of phonological features in basic vocabulary doesn't replicate when controlling for spatial and phylogenetic effects. The statistical over-representation of phonological features in the basic vocabulary of languages is often interpreted as reflecting potentially universal sound symbolic patterns. Ho...
cl_nlp
cs.CL
2025-12-08T13:24:53+00:00
2512.20487v1
Multi-temporal Adaptive Red-Green-Blue and Long-Wave Infrared Fusion for You Only Look Once-Based Landmine Detection from Unmanned Aerial Systems
Landmines remain a persistent humanitarian threat, with 110 million actively deployed mines across 60 countries, claiming 26,000 casualties annually. This research evaluates adaptive Red-Green-Blue (RGB) and Long-Wave Infrared (LWIR) fusion for Unmanned Aerial Systems (UAS)-based detection of surface-laid landmines, le...
Multi-temporal Adaptive Red-Green-Blue and Long-Wave Infrared Fusion for You Only Look Once-Based Landmine Detection from Unmanned Aerial Systems. Landmines remain a persistent humanitarian threat, with 110 million actively deployed mines across 60 countries, claiming 26,000 casualties annually. This research evaluates...
cv_vision
cs.CV
2025-12-23T16:26:47+00:00
2512.06590v1
Towards Efficient Hypergraph and Multi-LLM Agent Recommender Systems
Recommender Systems (RSs) have become the cornerstone of various applications such as e-commerce and social media platforms. The evolution of RSs is paramount in the digital era, in which personalised user experience is tailored to the user's preferences. Large Language Models (LLMs) have sparked a new paradigm - gener...
Towards Efficient Hypergraph and Multi-LLM Agent Recommender Systems. Recommender Systems (RSs) have become the cornerstone of various applications such as e-commerce and social media platforms. The evolution of RSs is paramount in the digital era, in which personalised user experience is tailored to the user's prefere...
ir_retrieval
cs.IR
2025-12-06T23:04:49+00:00
2512.05959v2
M4-RAG: A Massive-Scale Multilingual Multi-Cultural Multimodal RAG
Vision-language models (VLMs) have achieved strong performance in visual question answering (VQA), yet they remain constrained by static training data. Retrieval-Augmented Generation (RAG) mitigates this limitation by enabling access to up-to-date, culturally grounded, and multilingual information; however, multilingua...
M4-RAG: A Massive-Scale Multilingual Multi-Cultural Multimodal RAG. Vision-language models (VLMs) have achieved strong performance in visual question answering (VQA), yet they remain constrained by static training data. Retrieval-Augmented Generation (RAG) mitigates this limitation by enabling access to up-to-date, cul...
cl_nlp
cs.CL
2025-12-05T18:55:58+00:00
2512.23049v1
Accelerating Language Model Workflows with Prompt Choreography
Large language models are increasingly deployed in multi-agent workflows. We introduce Prompt Choreography, a framework that efficiently executes LLM workflows by maintaining a dynamic, global KV cache. Each LLM call can attend to an arbitrary, reordered subset of previously encoded messages. Parallel calls are support...
Accelerating Language Model Workflows with Prompt Choreography. Large language models are increasingly deployed in multi-agent workflows. We introduce Prompt Choreography, a framework that efficiently executes LLM workflows by maintaining a dynamic, global KV cache. Each LLM call can attend to an arbitrary, reordered s...
cl_nlp
cs.CL
2025-12-28T19:21:11+00:00
2512.22396v1
HalluMat: Detecting Hallucinations in LLM-Generated Materials Science Content Through Multi-Stage Verification
Artificial Intelligence (AI), particularly Large Language Models (LLMs), is transforming scientific discovery, enabling rapid knowledge generation and hypothesis formulation. However, a critical challenge is hallucination, where LLMs generate factually incorrect or misleading information, compromising research integrit...
HalluMat: Detecting Hallucinations in LLM-Generated Materials Science Content Through Multi-Stage Verification. Artificial Intelligence (AI), particularly Large Language Models (LLMs), is transforming scientific discovery, enabling rapid knowledge generation and hypothesis formulation. However, a critical challenge is ...
ai_general
cs.AI
2025-12-26T22:16:12+00:00
2510.10981v2
In-Context Learning Is Provably Bayesian Inference: A Generalization Theory for Meta-Learning
This paper develops a finite-sample statistical theory for in-context learning (ICL), analyzed within a meta-learning framework that accommodates mixtures of diverse task types. We introduce a principled risk decomposition that separates the total ICL risk into two orthogonal components: Bayes Gap and Posterior Varianc...
In-Context Learning Is Provably Bayesian Inference: A Generalization Theory for Meta-Learning. This paper develops a finite-sample statistical theory for in-context learning (ICL), analyzed within a meta-learning framework that accommodates mixtures of diverse task types. We introduce a principled risk decomposition th...
stat_ml
stat.ML
2025-10-13T03:42:31+00:00
2509.07860v1
KLIPA: A Knowledge Graph and LLM-Driven QA Framework for IP Analysis
Effectively managing intellectual property is a significant challenge. Traditional methods for patent analysis depend on labor-intensive manual searches and rigid keyword matching. These approaches are often inefficient and struggle to reveal the complex relationships hidden within large patent datasets, hindering stra...
KLIPA: A Knowledge Graph and LLM-Driven QA Framework for IP Analysis. Effectively managing intellectual property is a significant challenge. Traditional methods for patent analysis depend on labor-intensive manual searches and rigid keyword matching. These approaches are often inefficient and struggle to reveal the com...
ir_retrieval
cs.IR
2025-09-09T15:40:23+00:00
2512.11567v1
Extending a Parliamentary Corpus with MPs' Tweets: Automatic Annotation and Evaluation Using MultiParTweet
Social media serves as a critical medium in modern politics because it both reflects politicians' ideologies and facilitates communication with younger generations. We present MultiParTweet, a multilingual tweet corpus from X that connects politicians' social media discourse with German political corpus GerParCor, ther...
Extending a Parliamentary Corpus with MPs' Tweets: Automatic Annotation and Evaluation Using MultiParTweet. Social media serves as a critical medium in modern politics because it both reflects politicians' ideologies and facilitates communication with younger generations. We present MultiParTweet, a multilingual tweet ...
cl_nlp
cs.CL
2025-12-12T13:55:02+00:00
2512.18317v1
Trustworthy and Explainable Deep Reinforcement Learning for Safe and Energy-Efficient Process Control: A Use Case in Industrial Compressed Air Systems
This paper presents a trustworthy reinforcement learning approach for the control of industrial compressed air systems. We develop a framework that enables safe and energy-efficient operation under realistic boundary conditions and introduce a multi-level explainability pipeline combining input perturbation tests, grad...
Trustworthy and Explainable Deep Reinforcement Learning for Safe and Energy-Efficient Process Control: A Use Case in Industrial Compressed Air Systems. This paper presents a trustworthy reinforcement learning approach for the control of industrial compressed air systems. We develop a framework that enables safe and ene...
lg_ml
cs.LG
2025-12-20T11:11:49+00:00
2601.03262v1
Roles of MLLMs in Visually Rich Document Retrieval for RAG: A Survey
Visually rich documents (VRDs) challenge retrieval-augmented generation (RAG) with layout-dependent semantics, brittle OCR, and evidence spread across complex figures and structured tables. This survey examines how Multimodal Large Language Models (MLLMs) are being used to make VRD retrieval practical for RAG. We organ...
Roles of MLLMs in Visually Rich Document Retrieval for RAG: A Survey. Visually rich documents (VRDs) challenge retrieval-augmented generation (RAG) with layout-dependent semantics, brittle OCR, and evidence spread across complex figures and structured tables. This survey examines how Multimodal Large Language Models (M...
ir_retrieval
cs.IR
2025-12-16T16:32:10+00:00
2512.11835v1
A Monad-Based Clause Architecture for Artificial Age Score (AAS) in Large Language Models
Large language models (LLMs) are often deployed as powerful yet opaque systems, leaving open how their internal memory and "self-like" behavior should be governed in a principled and auditable way. The Artificial Age Score (AAS) was previously introduced and mathematically justified through three theorems that characte...
A Monad-Based Clause Architecture for Artificial Age Score (AAS) in Large Language Models. Large language models (LLMs) are often deployed as powerful yet opaque systems, leaving open how their internal memory and "self-like" behavior should be governed in a principled and auditable way. The Artificial Age Score (AAS) ...
ai_general
cs.AI
2025-12-03T12:48:40+00:00
2509.00200v2
Simulation-based inference of yeast centromeres
The chromatin folding and the spatial arrangement of chromosomes in the cell play a crucial role in DNA replication and genes expression. An improper chromatin folding could lead to malfunctions and, over time, diseases. For eukaryotes, centromeres are essential for proper chromosome segregation and folding. Despite ex...
Simulation-based inference of yeast centromeres. The chromatin folding and the spatial arrangement of chromosomes in the cell play a crucial role in DNA replication and genes expression. An improper chromatin folding could lead to malfunctions and, over time, diseases. For eukaryotes, centromeres are essential for prop...
stat_ml
stat.ML
2025-08-29T19:23:06+00:00
2512.08273v1
AgentEval: Generative Agents as Reliable Proxies for Human Evaluation of AI-Generated Content
Modern businesses are increasingly challenged by the time and expense required to generate and assess high-quality content. Human writers face time constraints, and extrinsic evaluations can be costly. While Large Language Models (LLMs) offer potential in content creation, concerns about the quality of AI-generated con...
AgentEval: Generative Agents as Reliable Proxies for Human Evaluation of AI-Generated Content. Modern businesses are increasingly challenged by the time and expense required to generate and assess high-quality content. Human writers face time constraints, and extrinsic evaluations can be costly. While Large Language Mo...
ai_general
cs.AI
2025-12-09T06:03:25+00:00
2512.15691v1
Multi-Modal Semantic Communication
Semantic communication aims to transmit information most relevant to a task rather than raw data, offering significant gains in communication efficiency for applications such as telepresence, augmented reality, and remote sensing. Recent transformer-based approaches have used self-attention maps to identify informative...
Multi-Modal Semantic Communication. Semantic communication aims to transmit information most relevant to a task rather than raw data, offering significant gains in communication efficiency for applications such as telepresence, augmented reality, and remote sensing. Recent transformer-based approaches have used self-at...
lg_ml
cs.LG
2025-12-17T18:47:22+00:00
2512.16188v1
A Multi-scale Fused Graph Neural Network with Inter-view Contrastive Learning for Spatial Transcriptomics Data Clustering
Spatial transcriptomics enables genome-wide expression analysis within native tissue context, yet identifying spatial domains remains challenging due to complex gene-spatial interactions. Existing methods typically process spatial and feature views separately, fusing only at output level - an "encode-separately, fuse-l...
A Multi-scale Fused Graph Neural Network with Inter-view Contrastive Learning for Spatial Transcriptomics Data Clustering. Spatial transcriptomics enables genome-wide expression analysis within native tissue context, yet identifying spatial domains remains challenging due to complex gene-spatial interactions. Existing ...
lg_ml
cs.LG
2025-12-18T05:13:55+00:00
2512.10687v2
Safe for Whom? Rethinking How We Evaluate the Safety of LLMs for Real Users
Safety evaluations of large language models (LLMs) typically focus on universal risks like dangerous capabilities or undesirable propensities. However, millions use LLMs for personal advice on high-stakes topics like finance and health, where harms are context-dependent rather than universal. While frameworks like the ...
Safe for Whom? Rethinking How We Evaluate the Safety of LLMs for Real Users. Safety evaluations of large language models (LLMs) typically focus on universal risks like dangerous capabilities or undesirable propensities. However, millions use LLMs for personal advice on high-stakes topics like finance and health, where ...
ai_general
cs.AI
2025-12-11T14:34:40+00:00
2512.17593v3
A Unified Representation of Neural Networks Architectures
In this paper we consider the limiting case of neural networks (NNs) architectures when the number of neurons in each hidden layer and the number of hidden layers tend to infinity thus forming a continuum, and we derive approximation errors as a function of the number of neurons and/or hidden layers. Firstly, we consid...
A Unified Representation of Neural Networks Architectures. In this paper we consider the limiting case of neural networks (NNs) architectures when the number of neurons in each hidden layer and the number of hidden layers tend to infinity thus forming a continuum, and we derive approximation errors as a function of the...
lg_ml
cs.LG
2025-12-19T14:01:50+00:00
2512.20140v1
Enhancing Zero-Shot Time Series Forecasting in Off-the-Shelf LLMs via Noise Injection
Large Language Models (LLMs) have demonstrated effectiveness as zero-shot time series (TS) forecasters. The key challenge lies in tokenizing TS data into textual representations that align with LLMs' pre-trained knowledge. While existing work often relies on fine-tuning specialized modules to bridge this gap, a distinc...
Enhancing Zero-Shot Time Series Forecasting in Off-the-Shelf LLMs via Noise Injection. Large Language Models (LLMs) have demonstrated effectiveness as zero-shot time series (TS) forecasters. The key challenge lies in tokenizing TS data into textual representations that align with LLMs' pre-trained knowledge. While exis...
ai_general
cs.AI
2025-12-23T08:02:33+00:00
2512.18934v1
When Less is More: 8-bit Quantization Improves Continual Learning in Large Language Models
Catastrophic forgetting poses a fundamental challenge in continual learning, particularly when models are quantized for deployment efficiency. We systematically investigate the interplay between quantization precision (FP16, INT8, INT4) and replay buffer strategies in large language models, revealing unexpected dynamic...
When Less is More: 8-bit Quantization Improves Continual Learning in Large Language Models. Catastrophic forgetting poses a fundamental challenge in continual learning, particularly when models are quantized for deployment efficiency. We systematically investigate the interplay between quantization precision (FP16, INT...
lg_ml
cs.LG
2025-12-22T00:51:39+00:00
2511.13025v1
Reconstruction of Manifold Distances from Noisy Observations
We consider the problem of reconstructing the intrinsic geometry of a manifold from noisy pairwise distance observations. Specifically, let $M$ denote a diameter 1 d-dimensional manifold and $μ$ a probability measure on $M$ that is mutually absolutely continuous with the volume measure. Suppose $X_1,\dots,X_N$ are i.i....
Reconstruction of Manifold Distances from Noisy Observations. We consider the problem of reconstructing the intrinsic geometry of a manifold from noisy pairwise distance observations. Specifically, let $M$ denote a diameter 1 d-dimensional manifold and $μ$ a probability measure on $M$ that is mutually absolutely contin...
stat_ml
stat.ML
2025-11-17T06:24:31+00:00
2512.18815v1
Controllable Probabilistic Forecasting with Stochastic Decomposition Layers
AI weather prediction ensembles with latent noise injection and optimized with the continuous ranked probability score (CRPS) have produced both accurate and well-calibrated predictions with far less computational cost compared with diffusion-based methods. However, current CRPS ensemble approaches vary in their traini...
Controllable Probabilistic Forecasting with Stochastic Decomposition Layers. AI weather prediction ensembles with latent noise injection and optimized with the continuous ranked probability score (CRPS) have produced both accurate and well-calibrated predictions with far less computational cost compared with diffusion-...
lg_ml
cs.LG
2025-12-21T17:10:00+00:00
2512.24157v1
Training Report of TeleChat3-MoE
TeleChat3-MoE is the latest series of TeleChat large language models, featuring a Mixture-of-Experts (MoE) architecture with parameter counts ranging from 105 billion to over one trillion,trained end-to-end on Ascend NPU cluster. This technical report mainly presents the underlying training infrastructure that enables ...
Training Report of TeleChat3-MoE. TeleChat3-MoE is the latest series of TeleChat large language models, featuring a Mixture-of-Experts (MoE) architecture with parameter counts ranging from 105 billion to over one trillion,trained end-to-end on Ascend NPU cluster. This technical report mainly presents the underlying tra...
cl_nlp
cs.CL
2025-12-30T11:42:14+00:00
2512.21220v2
RoboSafe: Safeguarding Embodied Agents via Executable Safety Logic
Embodied agents powered by vision-language models (VLMs) are increasingly capable of executing complex real-world tasks, yet they remain vulnerable to hazardous instructions that may trigger unsafe behaviors. Runtime safety guardrails, which intercept hazardous actions during task execution, offer a promising solution ...
RoboSafe: Safeguarding Embodied Agents via Executable Safety Logic. Embodied agents powered by vision-language models (VLMs) are increasingly capable of executing complex real-world tasks, yet they remain vulnerable to hazardous instructions that may trigger unsafe behaviors. Runtime safety guardrails, which intercept ...
ai_general
cs.AI
2025-12-24T15:01:26+00:00
2512.18834v2
Mix, MinHash, and Match: Cross-Source Agreement for Multilingual Pretraining Datasets
Multilingual data from the web is essential for LLM pretraining. Yet, scraping it is expensive, and research groups repeatedly crawl the same content. For example, we found that over 40\% of tokens across major Arabic web corpora are duplicated between sources. In this work, we propose to use this wasteful redundancy a...
Mix, MinHash, and Match: Cross-Source Agreement for Multilingual Pretraining Datasets. Multilingual data from the web is essential for LLM pretraining. Yet, scraping it is expensive, and research groups repeatedly crawl the same content. For example, we found that over 40\% of tokens across major Arabic web corpora are...
cl_nlp
cs.CL
2025-12-21T17:36:26+00:00
2512.21613v1
AMS-IO-Bench and AMS-IO-Agent: Benchmarking and Structured Reasoning for Analog and Mixed-Signal Integrated Circuit Input/Output Design
In this paper, we propose AMS-IO-Agent, a domain-specialized LLM-based agent for structure-aware input/output (I/O) subsystem generation in analog and mixed-signal (AMS) integrated circuits (ICs). The central contribution of this work is a framework that connects natural language design intent with industrial-level AMS...
AMS-IO-Bench and AMS-IO-Agent: Benchmarking and Structured Reasoning for Analog and Mixed-Signal Integrated Circuit Input/Output Design. In this paper, we propose AMS-IO-Agent, a domain-specialized LLM-based agent for structure-aware input/output (I/O) subsystem generation in analog and mixed-signal (AMS) integrated ci...
ai_general
cs.AI
2025-12-25T10:20:55+00:00
2512.16317v1
Design and Evaluation of Cost-Aware PoQ for Decentralized LLM Inference
Decentralized large language model (LLM) inference promises transparent and censorship resistant access to advanced AI, yet existing verification approaches struggle to scale to modern models. Proof of Quality (PoQ) replaces cryptographic verification of computation with consensus over output quality, but the original ...
Design and Evaluation of Cost-Aware PoQ for Decentralized LLM Inference. Decentralized large language model (LLM) inference promises transparent and censorship resistant access to advanced AI, yet existing verification approaches struggle to scale to modern models. Proof of Quality (PoQ) replaces cryptographic verifica...
ai_general
cs.AI
2025-12-18T08:57:17+00:00
2512.20661v1
From Fake Focus to Real Precision: Confusion-Driven Adversarial Attention Learning in Transformers
Transformer-based models have been widely adopted for sentiment analysis tasks due to their exceptional ability to capture contextual information. However, these methods often exhibit suboptimal accuracy in certain scenarios. By analyzing their attention distributions, we observe that existing models tend to allocate a...
From Fake Focus to Real Precision: Confusion-Driven Adversarial Attention Learning in Transformers. Transformer-based models have been widely adopted for sentiment analysis tasks due to their exceptional ability to capture contextual information. However, these methods often exhibit suboptimal accuracy in certain scena...
ai_general
cs.AI
2025-12-19T01:48:25+00:00
2603.15622v1
SAC-NeRF: Adaptive Ray Sampling for Neural Radiance Fields via Soft Actor-Critic Reinforcement Learning
Neural Radiance Fields (NeRF) have achieved photorealistic novel view synthesis but suffer from computational inefficiency due to dense ray sampling during volume rendering. We propose SAC-NeRF, a reinforcement learning framework that learns adaptive sampling policies using Soft Actor-Critic (SAC). Our method formulate...
SAC-NeRF: Adaptive Ray Sampling for Neural Radiance Fields via Soft Actor-Critic Reinforcement Learning. Neural Radiance Fields (NeRF) have achieved photorealistic novel view synthesis but suffer from computational inefficiency due to dense ray sampling during volume rendering. We propose SAC-NeRF, a reinforcement lear...
cv_vision
cs.CV
2025-12-31T08:08:55+00:00
2512.21040v1
A Large-Depth-Range Layer-Based Hologram Dataset for Machine Learning-Based 3D Computer-Generated Holography
Machine learning-based computer-generated holography (ML-CGH) has advanced rapidly in recent years, yet progress is constrained by the limited availability of high-quality, large-scale hologram datasets. To address this, we present KOREATECH-CGH, a publicly available dataset comprising 6,000 pairs of RGB-D images and c...
A Large-Depth-Range Layer-Based Hologram Dataset for Machine Learning-Based 3D Computer-Generated Holography. Machine learning-based computer-generated holography (ML-CGH) has advanced rapidly in recent years, yet progress is constrained by the limited availability of high-quality, large-scale hologram datasets. To add...
cv_vision
cs.CV
2025-12-24T08:07:39+00:00
2512.22886v1
Theory and Algorithms for Learning with Multi-Class Abstention and Multi-Expert Deferral
Large language models (LLMs) have achieved remarkable performance but face critical challenges: hallucinations and high inference costs. Leveraging multiple experts offers a solution: deferring uncertain inputs to more capable experts improves reliability, while routing simpler queries to smaller, distilled models enha...
Theory and Algorithms for Learning with Multi-Class Abstention and Multi-Expert Deferral. Large language models (LLMs) have achieved remarkable performance but face critical challenges: hallucinations and high inference costs. Leveraging multiple experts offers a solution: deferring uncertain inputs to more capable exp...
lg_ml
cs.LG
2025-12-28T11:33:39+00:00
2512.24111v1
Guided Diffusion-based Generation of Adversarial Objects for Real-World Monocular Depth Estimation Attacks
Monocular Depth Estimation (MDE) serves as a core perception module in autonomous driving systems, but it remains highly susceptible to adversarial attacks. Errors in depth estimation may propagate through downstream decision making and influence overall traffic safety. Existing physical attacks primarily rely on textu...
Guided Diffusion-based Generation of Adversarial Objects for Real-World Monocular Depth Estimation Attacks. Monocular Depth Estimation (MDE) serves as a core perception module in autonomous driving systems, but it remains highly susceptible to adversarial attacks. Errors in depth estimation may propagate through downst...
cv_vision
cs.CV
2025-12-30T09:41:41+00:00
2512.20513v1
Recurrent Off-Policy Deep Reinforcement Learning Doesn't Have to be Slow
Recurrent off-policy deep reinforcement learning models achieve state-of-the-art performance but are often sidelined due to their high computational demands. In response, we introduce RISE (Recurrent Integration via Simplified Encodings), a novel approach that can leverage recurrent networks in any image-based off-poli...
Recurrent Off-Policy Deep Reinforcement Learning Doesn't Have to be Slow. Recurrent off-policy deep reinforcement learning models achieve state-of-the-art performance but are often sidelined due to their high computational demands. In response, we introduce RISE (Recurrent Integration via Simplified Encodings), a novel...
lg_ml
cs.LG
2025-12-23T17:02:17+00:00
2512.22178v1
Wireless Traffic Prediction with Large Language Model
The growing demand for intelligent, adaptive resource management in next-generation wireless networks has underscored the importance of accurate and scalable wireless traffic prediction. While recent advancements in deep learning and foundation models such as large language models (LLMs) have demonstrated promising for...
Wireless Traffic Prediction with Large Language Model. The growing demand for intelligent, adaptive resource management in next-generation wireless networks has underscored the importance of accurate and scalable wireless traffic prediction. While recent advancements in deep learning and foundation models such as large...
lg_ml
cs.LG
2025-12-19T04:47:40+00:00
2509.07543v2
Asynchronous Gossip Algorithms for Rank-Based Statistical Methods
As decentralized AI and edge intelligence become increasingly prevalent, ensuring robustness and trustworthiness in such distributed settings has become a critical issue-especially in the presence of corrupted or adversarial data. Traditional decentralized algorithms are vulnerable to data contamination as they typical...
Asynchronous Gossip Algorithms for Rank-Based Statistical Methods. As decentralized AI and edge intelligence become increasingly prevalent, ensuring robustness and trustworthiness in such distributed settings has become a critical issue-especially in the presence of corrupted or adversarial data. Traditional decentrali...
stat_ml
stat.ML
2025-09-09T09:23:36+00:00
2512.20985v1
A Blockchain-Monitored Agentic AI Architecture for Trusted Perception-Reasoning-Action Pipelines
The application of agentic AI systems in autonomous decision-making is growing in the areas of healthcare, smart cities, digital forensics, and supply chain management. Even though these systems are flexible and offer real-time reasoning, they also raise concerns of trust and oversight, and integrity of the information...
A Blockchain-Monitored Agentic AI Architecture for Trusted Perception-Reasoning-Action Pipelines. The application of agentic AI systems in autonomous decision-making is growing in the areas of healthcare, smart cities, digital forensics, and supply chain management. Even though these systems are flexible and offer real...
ai_general
cs.AI
2025-12-24T06:20:28+00:00
2512.22800v1
Medical Scene Reconstruction and Segmentation based on 3D Gaussian Representation
3D reconstruction of medical images is a key technology in medical image analysis and clinical diagnosis, providing structural visualization support for disease assessment and surgical planning. Traditional methods are computationally expensive and prone to structural discontinuities and loss of detail in sparse slices...
Medical Scene Reconstruction and Segmentation based on 3D Gaussian Representation. 3D reconstruction of medical images is a key technology in medical image analysis and clinical diagnosis, providing structural visualization support for disease assessment and surgical planning. Traditional methods are computationally ex...
cv_vision
cs.CV
2025-12-28T06:18:11+00:00
2512.16125v1
Convolutional Lie Operator for Sentence Classification
Traditional Convolutional Neural Networks have been successful in capturing local, position-invariant features in text, but their capacity to model complex transformation within language can be further explored. In this work, we explore a novel approach by integrating Lie Convolutions into Convolutional-based sentence ...
Convolutional Lie Operator for Sentence Classification. Traditional Convolutional Neural Networks have been successful in capturing local, position-invariant features in text, but their capacity to model complex transformation within language can be further explored. In this work, we explore a novel approach by integra...
cl_nlp
cs.CL
2025-12-18T03:23:37+00:00
2512.08270v1
Reasoning Models Ace the CFA Exams
Previous research has reported that large language models (LLMs) demonstrate poor performance on the Chartered Financial Analyst (CFA) exams. However, recent reasoning models have achieved strong results on graduate-level academic and professional examinations across various disciplines. In this paper, we evaluate stat...
Reasoning Models Ace the CFA Exams. Previous research has reported that large language models (LLMs) demonstrate poor performance on the Chartered Financial Analyst (CFA) exams. However, recent reasoning models have achieved strong results on graduate-level academic and professional examinations across various discipli...
ai_general
cs.AI
2025-12-09T05:57:19+00:00
2512.10996v1
MedBioRAG: Semantic Search and Retrieval-Augmented Generation with Large Language Models for Medical and Biological QA
Recent advancements in retrieval-augmented generation (RAG) have significantly enhanced the ability of large language models (LLMs) to perform complex question-answering (QA) tasks. In this paper, we introduce MedBioRAG, a retrieval-augmented model designed to improve biomedical QA performance through a combination of ...
MedBioRAG: Semantic Search and Retrieval-Augmented Generation with Large Language Models for Medical and Biological QA. Recent advancements in retrieval-augmented generation (RAG) have significantly enhanced the ability of large language models (LLMs) to perform complex question-answering (QA) tasks. In this paper, we ...
cl_nlp
cs.CL
2025-12-10T15:43:25+00:00
2512.15358v1
Dual-Density Inference for Efficient Language Model Reasoning
Large Language Models (LLMs) have shown impressive capabilities in complex reasoning tasks. However, current approaches employ uniform language density for both intermediate reasoning and final answers, leading to computational inefficiency. Our observation found that reasoning process serves a computational function f...
Dual-Density Inference for Efficient Language Model Reasoning. Large Language Models (LLMs) have shown impressive capabilities in complex reasoning tasks. However, current approaches employ uniform language density for both intermediate reasoning and final answers, leading to computational inefficiency. Our observation...
cl_nlp
cs.CL
2025-12-17T12:04:05+00:00
2512.18947v1
Clustering-based Transfer Learning for Dynamic Multimodal MultiObjective Evolutionary Algorithm
Dynamic multimodal multiobjective optimization presents the dual challenge of simultaneously tracking multiple equivalent pareto optimal sets and maintaining population diversity in time-varying environments. However, existing dynamic multiobjective evolutionary algorithms often neglect solution modality, whereas stati...
Clustering-based Transfer Learning for Dynamic Multimodal MultiObjective Evolutionary Algorithm. Dynamic multimodal multiobjective optimization presents the dual challenge of simultaneously tracking multiple equivalent pareto optimal sets and maintaining population diversity in time-varying environments. However, exist...
ai_general
cs.AI
2025-12-22T01:51:26+00:00
2512.09530v2
Transformers for Tabular Data: A Training Perspective of Self-Attention via Optimal Transport
This thesis examines self-attention training through the lens of Optimal Transport (OT) and develops an OT-based alternative for tabular classification. The study tracks intermediate projections of the self-attention layer during training and evaluates their evolution using discrete OT metrics, including Wasserstein di...
Transformers for Tabular Data: A Training Perspective of Self-Attention via Optimal Transport. This thesis examines self-attention training through the lens of Optimal Transport (OT) and develops an OT-based alternative for tabular classification. The study tracks intermediate projections of the self-attention layer du...
stat_ml
stat.ML
2025-12-10T11:11:52+00:00
2512.16086v1
Privacy Blur: Quantifying Privacy and Utility for Image Data Release
Image data collected in the wild often contains private information such as faces and license plates, and responsible data release must ensure that this information stays hidden. At the same time, released data should retain its usefulness for model-training. The standard method for private information obfuscation in i...
Privacy Blur: Quantifying Privacy and Utility for Image Data Release. Image data collected in the wild often contains private information such as faces and license plates, and responsible data release must ensure that this information stays hidden. At the same time, released data should retain its usefulness for model-...
lg_ml
cs.LG
2025-12-18T02:01:17+00:00
2512.07179v1
PICKT: Practical Interlinked Concept Knowledge Tracing for Personalized Learning using Knowledge Map Concept Relations
With the recent surge in personalized learning, Intelligent Tutoring Systems (ITS) that can accurately track students' individual knowledge states and provide tailored learning paths based on this information are in demand as an essential task. This paper focuses on the core technology of Knowledge Tracing (KT) models ...
PICKT: Practical Interlinked Concept Knowledge Tracing for Personalized Learning using Knowledge Map Concept Relations. With the recent surge in personalized learning, Intelligent Tutoring Systems (ITS) that can accurately track students' individual knowledge states and provide tailored learning paths based on this inf...
ai_general
cs.AI
2025-12-08T05:24:17+00:00
2512.14047v2
AsarRec: Adaptive Sequential Augmentation for Robust Self-supervised Sequential Recommendation
Sequential recommender systems have demonstrated strong capabilities in modeling users' dynamic preferences and capturing item transition patterns. However, real-world user behaviors are often noisy due to factors such as human errors, uncertainty, and behavioral ambiguity, which can lead to degraded recommendation per...
AsarRec: Adaptive Sequential Augmentation for Robust Self-supervised Sequential Recommendation. Sequential recommender systems have demonstrated strong capabilities in modeling users' dynamic preferences and capturing item transition patterns. However, real-world user behaviors are often noisy due to factors such as hu...
ir_retrieval
cs.IR
2025-12-16T03:29:11+00:00
2509.19209v1
A Knowledge Graph and a Tripartite Evaluation Framework Make Retrieval-Augmented Generation Scalable and Transparent
Large Language Models (LLMs) have significantly enhanced conversational Artificial Intelligence(AI) chatbots; however, domain-specific accuracy and the avoidance of factual inconsistencies remain pressing challenges, particularly for large datasets. Designing an effective chatbot with appropriate methods and evaluating...
A Knowledge Graph and a Tripartite Evaluation Framework Make Retrieval-Augmented Generation Scalable and Transparent. Large Language Models (LLMs) have significantly enhanced conversational Artificial Intelligence(AI) chatbots; however, domain-specific accuracy and the avoidance of factual inconsistencies remain pressi...
ir_retrieval
cs.IR
2025-09-23T16:29:22+00:00
2512.19354v1
ReasonCD: A Multimodal Reasoning Large Model for Implicit Change-of-Interest Semantic Mining
Remote sensing image change detection is one of the fundamental tasks in remote sensing intelligent interpretation. Its core objective is to identify changes within change regions of interest (CRoI). Current multimodal large models encode rich human semantic knowledge, which is utilized for guidance in tasks such as re...
ReasonCD: A Multimodal Reasoning Large Model for Implicit Change-of-Interest Semantic Mining. Remote sensing image change detection is one of the fundamental tasks in remote sensing intelligent interpretation. Its core objective is to identify changes within change regions of interest (CRoI). Current multimodal large m...
cv_vision
cs.CV
2025-12-22T12:54:26+00:00
2512.05658v2
Multilingual Medical Reasoning for Question Answering with Large Language Models
Large Language Models (LLMs) with reasoning capabilities have recently demonstrated strong potential in medical Question Answering (QA). Existing approaches are largely English-focused and primarily rely on distillation from general-purpose LLMs, raising concerns about the reliability of their medical knowledge. In thi...
Multilingual Medical Reasoning for Question Answering with Large Language Models. Large Language Models (LLMs) with reasoning capabilities have recently demonstrated strong potential in medical Question Answering (QA). Existing approaches are largely English-focused and primarily rely on distillation from general-purpo...
cl_nlp
cs.CL
2025-12-05T12:05:46+00:00
2510.12014v1
Embedding the Teacher: Distilling vLLM Preferences for Scalable Image Retrieval
Text--image retrieval is necessary for applications such as product recommendation. Embedding-based approaches like CLIP enable efficient large-scale retrieval via vector similarity search, but they are primarily trained on literal caption-like text--image pairs and often fail to capture abstract or persona-driven attr...
Embedding the Teacher: Distilling vLLM Preferences for Scalable Image Retrieval. Text--image retrieval is necessary for applications such as product recommendation. Embedding-based approaches like CLIP enable efficient large-scale retrieval via vector similarity search, but they are primarily trained on literal caption...
ir_retrieval
cs.IR
2025-10-13T23:30:07+00:00
2512.03208v1
Uncertainty Quantification for Large Language Model Reward Learning under Heterogeneous Human Feedback
We study estimation and statistical inference for reward models used in aligning large language models (LLMs). A key component of LLM alignment is reinforcement learning from human feedback (RLHF), where humans compare pairs of model-generated answers and their preferences are used to train a reward model. However, hum...
Uncertainty Quantification for Large Language Model Reward Learning under Heterogeneous Human Feedback. We study estimation and statistical inference for reward models used in aligning large language models (LLMs). A key component of LLM alignment is reinforcement learning from human feedback (RLHF), where humans compa...
stat_ml
stat.ML
2025-12-02T20:22:25+00:00
2512.18599v2
Restore-R1: Efficient Image Restoration Agents via Reinforcement Learning with Multimodal LLM Perceptual Feedback
Complex image restoration aims to recover high-quality images from inputs affected by multiple degradations such as blur, noise, rain, and compression artifacts. Recent restoration agents, powered by vision-language models and large language models, offer promising restoration capabilities but suffer from significant e...
Restore-R1: Efficient Image Restoration Agents via Reinforcement Learning with Multimodal LLM Perceptual Feedback. Complex image restoration aims to recover high-quality images from inputs affected by multiple degradations such as blur, noise, rain, and compression artifacts. Recent restoration agents, powered by visio...
cv_vision
cs.CV
2025-12-21T05:12:25+00:00
2512.22880v1
Fundamental Novel Consistency Theory: $H$-Consistency Bounds
In machine learning, the loss functions optimized during training often differ from the target loss that defines task performance due to computational intractability or lack of differentiability. We present an in-depth study of the target loss estimation error relative to the surrogate loss estimation error. Our analys...
Fundamental Novel Consistency Theory: $H$-Consistency Bounds. In machine learning, the loss functions optimized during training often differ from the target loss that defines task performance due to computational intractability or lack of differentiability. We present an in-depth study of the target loss estimation err...
lg_ml
cs.LG
2025-12-28T11:02:20+00:00
2512.09563v1
System Report for CCL25-Eval Task 10: Prompt-Driven Large Language Model Merge for Fine-Grained Chinese Hate Speech Detection
The proliferation of hate speech on Chinese social media poses urgent societal risks, yet traditional systems struggle to decode context-dependent rhetorical strategies and evolving slang. To bridge this gap, we propose a novel three-stage LLM-based framework: Prompt Engineering, Supervised Fine-tuning, and LLM Merging...
System Report for CCL25-Eval Task 10: Prompt-Driven Large Language Model Merge for Fine-Grained Chinese Hate Speech Detection. The proliferation of hate speech on Chinese social media poses urgent societal risks, yet traditional systems struggle to decode context-dependent rhetorical strategies and evolving slang. To b...
cl_nlp
cs.CL
2025-12-10T11:58:34+00:00
2601.00086v2
RIMRULE: Improving Tool-Using Language Agents via MDL-Guided Rule Learning
Large language models (LLMs) often struggle to use tools reliably in domain-specific settings, where APIs may be idiosyncratic, under-documented, or tailored to private workflows. This highlights the need for effective adaptation to task-specific tools. We propose RIMRULE, a neuro-symbolic approach for LLM adaptation b...
RIMRULE: Improving Tool-Using Language Agents via MDL-Guided Rule Learning. Large language models (LLMs) often struggle to use tools reliably in domain-specific settings, where APIs may be idiosyncratic, under-documented, or tailored to private workflows. This highlights the need for effective adaptation to task-specif...
cl_nlp
cs.CL
2025-12-31T19:40:10+00:00
2509.09114v1
Modality Alignment with Multi-scale Bilateral Attention for Multimodal Recommendation
Multimodal recommendation systems are increasingly becoming foundational technologies for e-commerce and content platforms, enabling personalized services by jointly modeling users' historical behaviors and the multimodal features of items (e.g., visual and textual). However, most existing methods rely on either static...
Modality Alignment with Multi-scale Bilateral Attention for Multimodal Recommendation. Multimodal recommendation systems are increasingly becoming foundational technologies for e-commerce and content platforms, enabling personalized services by jointly modeling users' historical behaviors and the multimodal features of...
ir_retrieval
cs.IR
2025-09-11T02:52:26+00:00
2512.16924v1
The World is Your Canvas: Painting Promptable Events with Reference Images, Trajectories, and Text
We present WorldCanvas, a framework for promptable world events that enables rich, user-directed simulation by combining text, trajectories, and reference images. Unlike text-only approaches and existing trajectory-controlled image-to-video methods, our multimodal approach combines trajectories -- encoding motion, timi...
The World is Your Canvas: Painting Promptable Events with Reference Images, Trajectories, and Text. We present WorldCanvas, a framework for promptable world events that enables rich, user-directed simulation by combining text, trajectories, and reference images. Unlike text-only approaches and existing trajectory-contr...
cv_vision
cs.CV
2025-12-18T18:59:59+00:00
2512.07314v1
M-STAR: Multi-Scale Spatiotemporal Autoregression for Human Mobility Modeling
Modeling human mobility is vital for extensive applications such as transportation planning and epidemic modeling. With the rise of the Artificial Intelligence Generated Content (AIGC) paradigm, recent works explore synthetic trajectory generation using autoregressive and diffusion models. While these methods show prom...
M-STAR: Multi-Scale Spatiotemporal Autoregression for Human Mobility Modeling. Modeling human mobility is vital for extensive applications such as transportation planning and epidemic modeling. With the rise of the Artificial Intelligence Generated Content (AIGC) paradigm, recent works explore synthetic trajectory gene...
ai_general
cs.AI
2025-12-08T08:57:55+00:00
2512.16626v1
Stackelberg Learning from Human Feedback: Preference Optimization as a Sequential Game
We introduce Stackelberg Learning from Human Feedback (SLHF), a new framework for preference optimization. SLHF frames the alignment problem as a sequential-move game between two policies: a Leader, which commits to an action, and a Follower, which responds conditionally on the Leader's action. This approach decomposes...
Stackelberg Learning from Human Feedback: Preference Optimization as a Sequential Game. We introduce Stackelberg Learning from Human Feedback (SLHF), a new framework for preference optimization. SLHF frames the alignment problem as a sequential-move game between two policies: a Leader, which commits to an action, and a...
lg_ml
cs.LG
2025-12-18T15:03:23+00:00
2601.04207v1
Ideology as a Problem: Lightweight Logit Steering for Annotator-Specific Alignment in Social Media Analysis
LLMs internally organize political ideology along low-dimensional structures that are partially, but not fully aligned with human ideological space. This misalignment is systematic, model specific, and measurable. We introduce a lightweight linear probe that both quantifies the misalignment and minimally corrects the o...
Ideology as a Problem: Lightweight Logit Steering for Annotator-Specific Alignment in Social Media Analysis. LLMs internally organize political ideology along low-dimensional structures that are partially, but not fully aligned with human ideological space. This misalignment is systematic, model specific, and measurabl...
cl_nlp
cs.CL
2025-12-08T14:07:44+00:00
2512.21677v1
Dictionary-Transform Generative Adversarial Networks
Generative adversarial networks (GANs) are widely used for distribution learning, yet their classical formulations remain theoretically fragile, with ill-posed objectives, unstable training dynamics, and limited interpretability. In this work, we introduce \emph{Dictionary-Transform Generative Adversarial Networks} (DT...
Dictionary-Transform Generative Adversarial Networks. Generative adversarial networks (GANs) are widely used for distribution learning, yet their classical formulations remain theoretically fragile, with ill-posed objectives, unstable training dynamics, and limited interpretability. In this work, we introduce \emph{Dic...
lg_ml
cs.LG
2025-12-25T13:41:14+00:00
2512.04442v1
TaskEval: Synthesised Evaluation for Foundation-Model Tasks
Hallucinations are a key concern when creating applications that rely on Foundation models (FMs). Understanding where and how these subtle failures occur in an application relies on evaluation methods known as \textit{evals}. Prior work focuses on defining new eval methods or benchmark datasets for specific tasks. Howe...
TaskEval: Synthesised Evaluation for Foundation-Model Tasks. Hallucinations are a key concern when creating applications that rely on Foundation models (FMs). Understanding where and how these subtle failures occur in an application relies on evaluation methods known as \textit{evals}. Prior work focuses on defining ne...
ai_general
cs.AI
2025-12-04T04:19:24+00:00
2512.24763v1
UniC-Lift: Unified 3D Instance Segmentation via Contrastive Learning
3D Gaussian Splatting (3DGS) and Neural Radiance Fields (NeRF) have advanced novel-view synthesis. Recent methods extend multi-view 2D segmentation to 3D, enabling instance/semantic segmentation for better scene understanding. A key challenge is the inconsistency of 2D instance labels across views, leading to poor 3D p...
UniC-Lift: Unified 3D Instance Segmentation via Contrastive Learning. 3D Gaussian Splatting (3DGS) and Neural Radiance Fields (NeRF) have advanced novel-view synthesis. Recent methods extend multi-view 2D segmentation to 3D, enabling instance/semantic segmentation for better scene understanding. A key challenge is the ...
cv_vision
cs.CV
2025-12-31T10:20:01+00:00
2510.13094v1
Gaussian Certified Unlearning in High Dimensions: A Hypothesis Testing Approach
Machine unlearning seeks to efficiently remove the influence of selected data while preserving generalization. Significant progress has been made in low dimensions $(p \ll n)$, but high dimensions pose serious theoretical challenges as standard optimization assumptions of $Ω(1)$ strong convexity and $O(1)$ smoothness o...
Gaussian Certified Unlearning in High Dimensions: A Hypothesis Testing Approach. Machine unlearning seeks to efficiently remove the influence of selected data while preserving generalization. Significant progress has been made in low dimensions $(p \ll n)$, but high dimensions pose serious theoretical challenges as sta...
stat_ml
stat.ML
2025-10-15T02:28:12+00:00
2512.19743v1
From Theory to Throughput: CUDA-Optimized APML for Large-Batch 3D Learning
Loss functions are fundamental to learning accurate 3D point cloud models, yet common choices trade geometric fidelity for computational cost. Chamfer Distance is efficient but permits many-to-one correspondences, while Earth Mover Distance better reflects one-to-one transport at high computational cost. APML approxima...
From Theory to Throughput: CUDA-Optimized APML for Large-Batch 3D Learning. Loss functions are fundamental to learning accurate 3D point cloud models, yet common choices trade geometric fidelity for computational cost. Chamfer Distance is efficient but permits many-to-one correspondences, while Earth Mover Distance bet...
lg_ml
cs.LG
2025-12-17T23:18:51+00:00
2512.00930v1
Thompson Sampling for Multi-Objective Linear Contextual Bandit
We study the multi-objective linear contextual bandit problem, where multiple possible conflicting objectives must be optimized simultaneously. We propose \texttt{MOL-TS}, the \textit{first} Thompson Sampling algorithm with Pareto regret guarantees for this problem. Unlike standard approaches that compute an empirical ...
Thompson Sampling for Multi-Objective Linear Contextual Bandit. We study the multi-objective linear contextual bandit problem, where multiple possible conflicting objectives must be optimized simultaneously. We propose \texttt{MOL-TS}, the \textit{first} Thompson Sampling algorithm with Pareto regret guarantees for thi...
stat_ml
stat.ML
2025-11-30T15:18:01+00:00
2512.21434v1
Scalable Deep Subspace Clustering Network
Subspace clustering methods face inherent scalability limits due to the $O(n^3)$ cost (with $n$ denoting the number of data samples) of constructing full $n\times n$ affinities and performing spectral decomposition. While deep learning-based approaches improve feature extraction, they maintain this computational bottle...
Scalable Deep Subspace Clustering Network. Subspace clustering methods face inherent scalability limits due to the $O(n^3)$ cost (with $n$ denoting the number of data samples) of constructing full $n\times n$ affinities and performing spectral decomposition. While deep learning-based approaches improve feature extracti...
cv_vision
cs.CV
2025-12-24T21:46:38+00:00
2512.24562v1
HaluNet: Multi-Granular Uncertainty Modeling for Efficient Hallucination Detection in LLM Question Answering
Large Language Models (LLMs) excel at question answering (QA) but often generate hallucinations, including factual errors or fabricated content. Detecting hallucinations from internal uncertainty signals is attractive due to its scalability and independence from external resources. Existing methods often aim to accurat...
HaluNet: Multi-Granular Uncertainty Modeling for Efficient Hallucination Detection in LLM Question Answering. Large Language Models (LLMs) excel at question answering (QA) but often generate hallucinations, including factual errors or fabricated content. Detecting hallucinations from internal uncertainty signals is att...
cl_nlp
cs.CL
2025-12-31T02:03:10+00:00
2512.19049v2
Decoupled Generative Modeling for Human-Object Interaction Synthesis
Synthesizing realistic human-object interaction (HOI) is essential for 3D computer vision and robotics, underpinning animation and embodied control. Existing approaches often require manually specified intermediate waypoints and place all optimization objectives on a single network, which increases complexity, reduces ...
Decoupled Generative Modeling for Human-Object Interaction Synthesis. Synthesizing realistic human-object interaction (HOI) is essential for 3D computer vision and robotics, underpinning animation and embodied control. Existing approaches often require manually specified intermediate waypoints and place all optimizatio...
cv_vision
cs.CV
2025-12-22T05:33:59+00:00
2512.22367v1
Subgoaling Relaxation-based Heuristics for Numeric Planning with Infinite Actions
Numeric planning with control parameters extends the standard numeric planning model by introducing action parameters as free numeric variables that must be instantiated during planning. This results in a potentially infinite number of applicable actions in a state. In this setting, off-the-shelf numeric heuristics tha...
Subgoaling Relaxation-based Heuristics for Numeric Planning with Infinite Actions. Numeric planning with control parameters extends the standard numeric planning model by introducing action parameters as free numeric variables that must be instantiated during planning. This results in a potentially infinite number of a...
ai_general
cs.AI
2025-12-26T20:05:57+00:00
2512.14563v1
Residual GRU+MHSA: A Lightweight Hybrid Recurrent Attention Model for Cardiovascular Disease Detection
Cardiovascular disease (CVD) remains the leading cause of mortality worldwide, underscoring the need for reliable and efficient predictive tools that support early intervention. Traditional diagnostic approaches rely on handcrafted features and clinician expertise, while machine learning methods improve reproducibility...
Residual GRU+MHSA: A Lightweight Hybrid Recurrent Attention Model for Cardiovascular Disease Detection. Cardiovascular disease (CVD) remains the leading cause of mortality worldwide, underscoring the need for reliable and efficient predictive tools that support early intervention. Traditional diagnostic approaches rely...
lg_ml
cs.LG
2025-12-16T16:33:59+00:00
2509.16446v1
Purely Semantic Indexing for LLM-based Generative Recommendation and Retrieval
Semantic identifiers (IDs) have proven effective in adapting large language models for generative recommendation and retrieval. However, existing methods often suffer from semantic ID conflicts, where semantically similar documents (or items) are assigned identical IDs. A common strategy to avoid conflicts is to append...
Purely Semantic Indexing for LLM-based Generative Recommendation and Retrieval. Semantic identifiers (IDs) have proven effective in adapting large language models for generative recommendation and retrieval. However, existing methods often suffer from semantic ID conflicts, where semantically similar documents (or item...
ir_retrieval
cs.IR
2025-09-19T21:59:55+00:00
2512.22473v5
Gradient Dynamics of Attention: How Cross-Entropy Sculpts Bayesian Manifolds
Transformers empirically perform precise probabilistic reasoning in carefully constructed ``Bayesian wind tunnels'' and in large-scale language models, yet the mechanisms by which gradient-based learning creates the required internal geometry remain opaque. We provide a complete first-order analysis of how cross-entrop...
Gradient Dynamics of Attention: How Cross-Entropy Sculpts Bayesian Manifolds. Transformers empirically perform precise probabilistic reasoning in carefully constructed ``Bayesian wind tunnels'' and in large-scale language models, yet the mechanisms by which gradient-based learning creates the required internal geometry...
stat_ml
stat.ML
2025-12-27T05:31:44+00:00
2512.22239v1
Multi-objective hybrid knowledge distillation for efficient deep learning in smart agriculture
Deploying deep learning models on resource-constrained edge devices remains a major challenge in smart agriculture due to the trade-off between computational efficiency and recognition accuracy. To address this challenge, this study proposes a hybrid knowledge distillation framework for developing a lightweight yet hig...
Multi-objective hybrid knowledge distillation for efficient deep learning in smart agriculture. Deploying deep learning models on resource-constrained edge devices remains a major challenge in smart agriculture due to the trade-off between computational efficiency and recognition accuracy. To address this challenge, th...
cv_vision
cs.CV
2025-12-23T15:33:55+00:00
2512.22069v1
Scaling Adversarial Training via Data Selection
Projected Gradient Descent (PGD) is a strong and widely used first-order adversarial attack, yet its computational cost scales poorly, as all training samples undergo identical iterative inner-loop optimization despite contributing unequally to robustness. Motivated by this inefficiency, we propose \emph{Selective Adve...
Scaling Adversarial Training via Data Selection. Projected Gradient Descent (PGD) is a strong and widely used first-order adversarial attack, yet its computational cost scales poorly, as all training samples undergo identical iterative inner-loop optimization despite contributing unequally to robustness. Motivated by t...
lg_ml
cs.LG
2025-12-26T15:50:33+00:00
2512.16171v1
Science Consultant Agent
The Science Consultant Agent is a web-based Artificial Intelligence (AI) tool that helps practitioners select and implement the most effective modeling strategy for AI-based solutions. It operates through four core components: Questionnaire, Smart Fill, Research-Guided Recommendation, and Prototype Builder. By combinin...
Science Consultant Agent. The Science Consultant Agent is a web-based Artificial Intelligence (AI) tool that helps practitioners select and implement the most effective modeling strategy for AI-based solutions. It operates through four core components: Questionnaire, Smart Fill, Research-Guided Recommendation, and Prot...
ai_general
cs.AI
2025-12-18T04:46:42+00:00
2512.04854v1
From Task Executors to Research Partners: Evaluating AI Co-Pilots Through Workflow Integration in Biomedical Research
Artificial intelligence systems are increasingly deployed in biomedical research. However, current evaluation frameworks may inadequately assess their effectiveness as research collaborators. This rapid review examines benchmarking practices for AI systems in preclinical biomedical research. Three major databases and t...
From Task Executors to Research Partners: Evaluating AI Co-Pilots Through Workflow Integration in Biomedical Research. Artificial intelligence systems are increasingly deployed in biomedical research. However, current evaluation frameworks may inadequately assess their effectiveness as research collaborators. This rapi...
ai_general
cs.AI
2025-12-04T14:37:46+00:00
2512.15684v1
High-Dimensional Partial Least Squares: Spectral Analysis and Fundamental Limitations
Partial Least Squares (PLS) is a widely used method for data integration, designed to extract latent components shared across paired high-dimensional datasets. Despite decades of practical success, a precise theoretical understanding of its behavior in high-dimensional regimes remains limited. In this paper, we study a...
High-Dimensional Partial Least Squares: Spectral Analysis and Fundamental Limitations. Partial Least Squares (PLS) is a widely used method for data integration, designed to extract latent components shared across paired high-dimensional datasets. Despite decades of practical success, a precise theoretical understanding...
stat_ml
stat.ML
2025-12-17T18:38:01+00:00
2510.21276v1
Pctx: Tokenizing Personalized Context for Generative Recommendation
Generative recommendation (GR) models tokenize each action into a few discrete tokens (called semantic IDs) and autoregressively generate the next tokens as predictions, showing advantages such as memory efficiency, scalability, and the potential to unify retrieval and ranking. Despite these benefits, existing tokeniza...
Pctx: Tokenizing Personalized Context for Generative Recommendation. Generative recommendation (GR) models tokenize each action into a few discrete tokens (called semantic IDs) and autoregressively generate the next tokens as predictions, showing advantages such as memory efficiency, scalability, and the potential to u...
ir_retrieval
cs.IR
2025-10-24T09:22:04+00:00