id
string
title
string
abstract
string
text
string
label
string
primary_category
string
published
string
2512.18437v1
MeniMV: A Multi-view Benchmark for Meniscus Injury Severity Grading
Precise grading of meniscal horn tears is critical in knee injury diagnosis but remains underexplored in automated MRI analysis. Existing methods often rely on coarse study-level labels or binary classification, lacking localization and severity information. In this paper, we introduce MeniMV, a multi-view benchmark da...
MeniMV: A Multi-view Benchmark for Meniscus Injury Severity Grading. Precise grading of meniscal horn tears is critical in knee injury diagnosis but remains underexplored in automated MRI analysis. Existing methods often rely on coarse study-level labels or binary classification, lacking localization and severity infor...
cv_vision
cs.CV
2025-12-20T17:22:55+00:00
2512.24289v1
Automated Analysis of Sustainability Reports: Using Large Language Models for the Extraction and Prediction of EU Taxonomy-Compliant KPIs
The manual, resource-intensive process of complying with the EU Taxonomy presents a significant challenge for companies. While Large Language Models (LLMs) offer a path to automation, research is hindered by a lack of public benchmark datasets. To address this gap, we introduce a novel, structured dataset from 190 corp...
Automated Analysis of Sustainability Reports: Using Large Language Models for the Extraction and Prediction of EU Taxonomy-Compliant KPIs. The manual, resource-intensive process of complying with the EU Taxonomy presents a significant challenge for companies. While Large Language Models (LLMs) offer a path to automatio...
cl_nlp
cs.CL
2025-12-30T15:28:03+00:00
2512.12804v1
Causal Counterfactuals Reconsidered
I develop a novel semantics for probabilities of counterfactuals that generalizes the standard Pearlian semantics: it applies to probabilistic causal models that cannot be extended into realistic structural causal models and are therefore beyond the scope of Pearl's semantics. This generalization is needed because, as ...
Causal Counterfactuals Reconsidered. I develop a novel semantics for probabilities of counterfactuals that generalizes the standard Pearlian semantics: it applies to probabilistic causal models that cannot be extended into realistic structural causal models and are therefore beyond the scope of Pearl's semantics. This ...
ai_general
cs.AI
2025-12-14T18:59:01+00:00
2512.15053v1
The Meta-Prompting Protocol: Orchestrating LLMs via Adversarial Feedback Loops
The transition of Large Language Models (LLMs) from stochastic chat interfaces to reliable software components necessitates a fundamental re-engineering of interaction paradigms. Current methodologies, predominantly heuristic-based "prompt engineering," fail to provide the deterministic guarantees required for mission-...
The Meta-Prompting Protocol: Orchestrating LLMs via Adversarial Feedback Loops. The transition of Large Language Models (LLMs) from stochastic chat interfaces to reliable software components necessitates a fundamental re-engineering of interaction paradigms. Current methodologies, predominantly heuristic-based "prompt ...
cl_nlp
cs.CL
2025-12-17T03:32:21+00:00
2511.20457v1
A Fully Probabilistic Tensor Network for Regularized Volterra System Identification
Modeling nonlinear systems with Volterra series is challenging because the number of kernel coefficients grows exponentially with the model order. This work introduces Bayesian Tensor Network Volterra kernel machines (BTN-V), extending the Bayesian Tensor Network framework to Volterra system identification. BTN-V repre...
A Fully Probabilistic Tensor Network for Regularized Volterra System Identification. Modeling nonlinear systems with Volterra series is challenging because the number of kernel coefficients grows exponentially with the model order. This work introduces Bayesian Tensor Network Volterra kernel machines (BTN-V), extending...
stat_ml
stat.ML
2025-11-25T16:24:52+00:00
2509.20239v1
Error Propagation in Dynamic Programming: From Stochastic Control to Option Pricing
This paper investigates theoretical and methodological foundations for stochastic optimal control (SOC) in discrete time. We start formulating the control problem in a general dynamic programming framework, introducing the mathematical structure needed for a detailed convergence analysis. The associate value function i...
Error Propagation in Dynamic Programming: From Stochastic Control to Option Pricing. This paper investigates theoretical and methodological foundations for stochastic optimal control (SOC) in discrete time. We start formulating the control problem in a general dynamic programming framework, introducing the mathematical...
stat_ml
stat.ML
2025-09-24T15:30:19+00:00
2512.22638v1
Likelihood-Preserving Embeddings for Statistical Inference
Modern machine learning embeddings provide powerful compression of high-dimensional data, yet they typically destroy the geometric structure required for classical likelihood-based statistical inference. This paper develops a rigorous theory of likelihood-preserving embeddings: learned representations that can replace ...
Likelihood-Preserving Embeddings for Statistical Inference. Modern machine learning embeddings provide powerful compression of high-dimensional data, yet they typically destroy the geometric structure required for classical likelihood-based statistical inference. This paper develops a rigorous theory of likelihood-pres...
stat_ml
stat.ML
2025-12-27T16:21:55+00:00
2512.22671v2
Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2
Structured width pruning of GLU-MLP layers, guided by the Maximum Absolute Weight (MAW) criterion, reveals a systematic dichotomy in how reducing the expansion ratio affects different model capabilities. While performance on tasks relying on parametric knowledge (e.g., MMLU, GSM8K) and perplexity metrics degrades predi...
Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2. Structured width pruning of GLU-MLP layers, guided by the Maximum Absolute Weight (MAW) criterion, reveals a systematic dichotomy in how reducing the expansion ratio affects different model capabilities. While performance on task...
cl_nlp
cs.CL
2025-12-27T18:09:57+00:00
2512.21521v1
First Provable Guarantees for Practical Private FL: Beyond Restrictive Assumptions
Federated Learning (FL) enables collaborative training on decentralized data. Differential privacy (DP) is crucial for FL, but current private methods often rely on unrealistic assumptions (e.g., bounded gradients or heterogeneity), hindering practical application. Existing works that relax these assumptions typically ...
First Provable Guarantees for Practical Private FL: Beyond Restrictive Assumptions. Federated Learning (FL) enables collaborative training on decentralized data. Differential privacy (DP) is crucial for FL, but current private methods often rely on unrealistic assumptions (e.g., bounded gradients or heterogeneity), hin...
lg_ml
cs.LG
2025-12-25T06:05:15+00:00
2511.19979v1
The 2nd Workshop on Human-Centered Recommender Systems
Recommender systems shape how people discover information, form opinions, and connect with society. Yet, as their influence grows, traditional metrics, e.g., accuracy, clicks, and engagement, no longer capture what truly matters to humans. The workshop on Human-Centered Recommender Systems (HCRS) calls for a paradigm s...
The 2nd Workshop on Human-Centered Recommender Systems. Recommender systems shape how people discover information, form opinions, and connect with society. Yet, as their influence grows, traditional metrics, e.g., accuracy, clicks, and engagement, no longer capture what truly matters to humans. The workshop on Human-Ce...
ir_retrieval
cs.IR
2025-11-25T06:49:14+00:00
2512.16245v1
AlignMerge - Alignment-Preserving Large Language Model Merging via Fisher-Guided Geometric Constraints
Merging large language models (LLMs) is a practical way to compose capabilities from multiple fine-tuned checkpoints without retraining. Yet standard schemes (linear weight soups, task vectors, and Fisher-weighted averaging) can preserve loss while quietly destroying alignment. We argue that merging is not a numerical ...
AlignMerge - Alignment-Preserving Large Language Model Merging via Fisher-Guided Geometric Constraints. Merging large language models (LLMs) is a practical way to compose capabilities from multiple fine-tuned checkpoints without retraining. Yet standard schemes (linear weight soups, task vectors, and Fisher-weighted av...
ai_general
cs.AI
2025-12-18T06:55:17+00:00
2512.20577v1
Improving ML Training Data with Gold-Standard Quality Metrics
Hand-tagged training data is essential to many machine learning tasks. However, training data quality control has received little attention in the literature, despite data quality varying considerably with the tagging exercise. We propose methods to evaluate and enhance the quality of hand-tagged training data using st...
Improving ML Training Data with Gold-Standard Quality Metrics. Hand-tagged training data is essential to many machine learning tasks. However, training data quality control has received little attention in the literature, despite data quality varying considerably with the tagging exercise. We propose methods to evaluat...
lg_ml
cs.LG
2025-12-23T18:21:24+00:00
2511.13418v1
Exploring Multi-Table Retrieval Through Iterative Search
Open-domain question answering over datalakes requires retrieving and composing information from multiple tables, a challenging subtask that demands semantic relevance and structural coherence (e.g., joinability). While exact optimization methods like Mixed-Integer Programming (MIP) can ensure coherence, their computat...
Exploring Multi-Table Retrieval Through Iterative Search. Open-domain question answering over datalakes requires retrieving and composing information from multiple tables, a challenging subtask that demands semantic relevance and structural coherence (e.g., joinability). While exact optimization methods like Mixed-Inte...
ir_retrieval
cs.IR
2025-11-17T14:31:33+00:00
2510.02067v2
Adaptive Kernel Selection for Stein Variational Gradient Descent
A central challenge in Bayesian inference is efficiently approximating posterior distributions. Stein Variational Gradient Descent (SVGD) is a popular variational inference method which transports a set of particles to approximate a target distribution. The SVGD dynamics are governed by a reproducing kernel Hilbert spa...
Adaptive Kernel Selection for Stein Variational Gradient Descent. A central challenge in Bayesian inference is efficiently approximating posterior distributions. Stein Variational Gradient Descent (SVGD) is a popular variational inference method which transports a set of particles to approximate a target distribution. ...
stat_ml
stat.ML
2025-10-02T14:33:57+00:00
2512.15586v2
Bolmo: Byteifying the Next Generation of Language Models
Recent advances in generative AI have been largely driven by large language models (LLMs), deep neural networks that operate over discrete units called tokens. To represent text, the vast majority of LLMs use words or word fragments as the tokens, known as subword tokenization. Subword tokenization obscures fine-graine...
Bolmo: Byteifying the Next Generation of Language Models. Recent advances in generative AI have been largely driven by large language models (LLMs), deep neural networks that operate over discrete units called tokens. To represent text, the vast majority of LLMs use words or word fragments as the tokens, known as subwo...
cl_nlp
cs.CL
2025-12-17T16:46:11+00:00
2512.22953v1
APO: Alpha-Divergence Preference Optimization
Two divergence regimes dominate modern alignment practice. Supervised fine-tuning and many distillation-style objectives implicitly minimize the forward KL divergence KL(q || pi_theta), yielding stable mode-covering updates but often under-exploiting high-reward modes. In contrast, PPO-style online reinforcement learni...
APO: Alpha-Divergence Preference Optimization. Two divergence regimes dominate modern alignment practice. Supervised fine-tuning and many distillation-style objectives implicitly minimize the forward KL divergence KL(q || pi_theta), yielding stable mode-covering updates but often under-exploiting high-reward modes. In ...
lg_ml
cs.LG
2025-12-28T14:51:03+00:00
2512.16795v1
From Facts to Conclusions : Integrating Deductive Reasoning in Retrieval-Augmented LLMs
Retrieval-Augmented Generation (RAG) grounds large language models (LLMs) in external evidence, but fails when retrieved sources conflict or contain outdated or subjective information. Prior work address these issues independently but lack unified reasoning supervision. We propose a reasoning-trace-augmented RAG framew...
From Facts to Conclusions : Integrating Deductive Reasoning in Retrieval-Augmented LLMs. Retrieval-Augmented Generation (RAG) grounds large language models (LLMs) in external evidence, but fails when retrieved sources conflict or contain outdated or subjective information. Prior work address these issues independently ...
cl_nlp
cs.CL
2025-12-18T17:27:51+00:00
2512.22976v2
A Context-Aware Temporal Modeling through Unified Multi-Scale Temporal Encoding and Hierarchical Sequence Learning for Single-Channel EEG Sleep Staging
Automatic sleep staging is a critical task in healthcare due to the global prevalence of sleep disorders. This study focuses on single-channel electroencephalography (EEG), a practical and widely available signal for automatic sleep staging. Existing approaches face challenges such as class imbalance, limited receptive...
A Context-Aware Temporal Modeling through Unified Multi-Scale Temporal Encoding and Hierarchical Sequence Learning for Single-Channel EEG Sleep Staging. Automatic sleep staging is a critical task in healthcare due to the global prevalence of sleep disorders. This study focuses on single-channel electroencephalography (...
lg_ml
cs.LG
2025-12-28T15:42:33+00:00
2512.15552v1
From Data to Dialogue: Unlocking Language for All
Traditional linguists have proposed the use of a General Service List (GSL) to assist new language learners in identifying the most important words in English. This process requires linguistic expertise, subjective input, and a considerable amount of time. We attempt to create our own GSL and evaluate its practicality ...
From Data to Dialogue: Unlocking Language for All. Traditional linguists have proposed the use of a General Service List (GSL) to assist new language learners in identifying the most important words in English. This process requires linguistic expertise, subjective input, and a considerable amount of time. We attempt t...
cl_nlp
cs.CL
2025-12-17T15:59:38+00:00
2512.17053v3
Knowledge Distillation with Structured Chain-of-Thought for Text-to-SQL
Deploying accurate Text-to-SQL systems at the enterprise level faces a difficult trilemma involving cost, security and performance. Current solutions force enterprises to choose between expensive, proprietary Large Language Models (LLMs) and low-performing Small Language Models (SLMs). Efforts to improve SLMs often rel...
Knowledge Distillation with Structured Chain-of-Thought for Text-to-SQL. Deploying accurate Text-to-SQL systems at the enterprise level faces a difficult trilemma involving cost, security and performance. Current solutions force enterprises to choose between expensive, proprietary Large Language Models (LLMs) and low-p...
cl_nlp
cs.CL
2025-12-18T20:41:22+00:00
2512.19539v1
StoryMem: Multi-shot Long Video Storytelling with Memory
Visual storytelling requires generating multi-shot videos with cinematic quality and long-range consistency. Inspired by human memory, we propose StoryMem, a paradigm that reformulates long-form video storytelling as iterative shot synthesis conditioned on explicit visual memory, transforming pre-trained single-shot vi...
StoryMem: Multi-shot Long Video Storytelling with Memory. Visual storytelling requires generating multi-shot videos with cinematic quality and long-range consistency. Inspired by human memory, we propose StoryMem, a paradigm that reformulates long-form video storytelling as iterative shot synthesis conditioned on expli...
cv_vision
cs.CV
2025-12-22T16:23:24+00:00
2512.13598v1
Textual Gradients are a Flawed Metaphor for Automatic Prompt Optimization
A well-engineered prompt can increase the performance of large language models; automatic prompt optimization techniques aim to increase performance without requiring human effort to tune the prompts. One leading class of prompt optimization techniques introduces the analogy of textual gradients. We investigate the beh...
Textual Gradients are a Flawed Metaphor for Automatic Prompt Optimization. A well-engineered prompt can increase the performance of large language models; automatic prompt optimization techniques aim to increase performance without requiring human effort to tune the prompts. One leading class of prompt optimization tec...
cl_nlp
cs.CL
2025-12-15T17:52:16+00:00
2512.20431v1
Skin Lesion Classification Using a Soft Voting Ensemble of Convolutional Neural Networks
Skin cancer can be identified by dermoscopic examination and ocular inspection, but early detection significantly increases survival chances. Artificial intelligence (AI), using annotated skin images and Convolutional Neural Networks (CNNs), improves diagnostic accuracy. This paper presents an early skin cancer classif...
Skin Lesion Classification Using a Soft Voting Ensemble of Convolutional Neural Networks. Skin cancer can be identified by dermoscopic examination and ocular inspection, but early detection significantly increases survival chances. Artificial intelligence (AI), using annotated skin images and Convolutional Neural Netwo...
cv_vision
cs.CV
2025-12-23T15:20:47+00:00
2511.07595v1
TurkEmbed4Retrieval: Turkish Embedding Model for Retrieval Task
In this work, we introduce TurkEmbed4Retrieval, a retrieval specialized variant of the TurkEmbed model originally designed for Natural Language Inference (NLI) and Semantic Textual Similarity (STS) tasks. By fine-tuning the base model on the MS MARCO TR dataset using advanced training techniques, including Matryoshka r...
TurkEmbed4Retrieval: Turkish Embedding Model for Retrieval Task. In this work, we introduce TurkEmbed4Retrieval, a retrieval specialized variant of the TurkEmbed model originally designed for Natural Language Inference (NLI) and Semantic Textual Similarity (STS) tasks. By fine-tuning the base model on the MS MARCO TR d...
ir_retrieval
cs.IR
2025-11-10T20:08:09+00:00
2512.16915v1
StereoPilot: Learning Unified and Efficient Stereo Conversion via Generative Priors
The rapid growth of stereoscopic displays, including VR headsets and 3D cinemas, has led to increasing demand for high-quality stereo video content. However, producing 3D videos remains costly and complex, while automatic Monocular-to-Stereo conversion is hindered by the limitations of the multi-stage ``Depth-Warp-Inpa...
StereoPilot: Learning Unified and Efficient Stereo Conversion via Generative Priors. The rapid growth of stereoscopic displays, including VR headsets and 3D cinemas, has led to increasing demand for high-quality stereo video content. However, producing 3D videos remains costly and complex, while automatic Monocular-to-...
cv_vision
cs.CV
2025-12-18T18:59:50+00:00
2510.23557v2
Minimizing Human Intervention in Online Classification
Training or fine-tuning large language model (LLM)-based systems often requires costly human feedback, yet there is limited understanding of how to minimize such intervention while maintaining strong error guarantees. We study this problem for LLM-based classification systems in an active learning framework: an agent s...
Minimizing Human Intervention in Online Classification. Training or fine-tuning large language model (LLM)-based systems often requires costly human feedback, yet there is limited understanding of how to minimize such intervention while maintaining strong error guarantees. We study this problem for LLM-based classifica...
stat_ml
stat.ML
2025-10-27T17:31:24+00:00
2512.17898v2
Humanlike AI Design Increases Anthropomorphism but Yields Divergent Outcomes on Engagement and Trust Globally
Over a billion users globally interact with AI systems engineered to mimic human traits. This development raises concerns that anthropomorphism, the attribution of human characteristics to AI, may foster over-reliance and misplaced trust. Yet, causal effects of humanlike AI design on users remain untested in ecological...
Humanlike AI Design Increases Anthropomorphism but Yields Divergent Outcomes on Engagement and Trust Globally. Over a billion users globally interact with AI systems engineered to mimic human traits. This development raises concerns that anthropomorphism, the attribution of human characteristics to AI, may foster over-...
ai_general
cs.AI
2025-12-19T18:57:53+00:00
2509.03131v1
RecBase: Generative Foundation Model Pretraining for Zero-Shot Recommendation
Recent advances in LLM-based recommendation have shown promise, yet their cross-domain generalization is hindered by a fundamental mismatch between language-centric pretraining and the recommendation task. Existing methods, relying on language-level knowledge, fail to capture dynamic, item-level user interests across d...
RecBase: Generative Foundation Model Pretraining for Zero-Shot Recommendation. Recent advances in LLM-based recommendation have shown promise, yet their cross-domain generalization is hindered by a fundamental mismatch between language-centric pretraining and the recommendation task. Existing methods, relying on langua...
ir_retrieval
cs.IR
2025-09-03T08:33:43+00:00
2512.16814v1
Grammar-Forced Translation of Natural Language to Temporal Logic using LLMs
Translating natural language (NL) into a formal language such as temporal logic (TL) is integral for human communication with robots and autonomous systems. State-of-the-art approaches decompose the task into a lifting of atomic propositions (APs) phase and a translation phase. However, existing methods struggle with a...
Grammar-Forced Translation of Natural Language to Temporal Logic using LLMs. Translating natural language (NL) into a formal language such as temporal logic (TL) is integral for human communication with robots and autonomous systems. State-of-the-art approaches decompose the task into a lifting of atomic propositions (...
cl_nlp
cs.CL
2025-12-18T17:55:15+00:00
2510.22049v3
Massive Memorization with Hundreds of Trillions of Parameters for Sequential Transducer Generative Recommenders
Modern large-scale recommendation systems rely heavily on user interaction history sequences to enhance the model performance. The advent of large language models and sequential modeling techniques, particularly transformer-like architectures, has led to significant advancements recently (e.g., HSTU, SIM, and TWIN mode...
Massive Memorization with Hundreds of Trillions of Parameters for Sequential Transducer Generative Recommenders. Modern large-scale recommendation systems rely heavily on user interaction history sequences to enhance the model performance. The advent of large language models and sequential modeling techniques, particul...
ir_retrieval
cs.IR
2025-10-24T22:17:49+00:00
2512.18129v2
TraCeR: Transformer-Based Competing Risk Analysis with Longitudinal Covariates
Survival analysis is a critical tool for modeling time-to-event data. Recent deep learning-based models have reduced various modeling assumptions including proportional hazard and linearity. However, a persistent challenge remains in incorporating longitudinal covariates, with prior work largely focusing on cross-secti...
TraCeR: Transformer-Based Competing Risk Analysis with Longitudinal Covariates. Survival analysis is a critical tool for modeling time-to-event data. Recent deep learning-based models have reduced various modeling assumptions including proportional hazard and linearity. However, a persistent challenge remains in incorp...
lg_ml
cs.LG
2025-12-19T23:24:47+00:00
2512.22760v1
Neighbor-Aware Token Reduction via Hilbert Curve for Vision Transformers
Vision Transformers (ViTs) have achieved remarkable success in visual recognition tasks, but redundant token representations limit their computational efficiency. Existing token merging and pruning strategies often overlook spatial continuity and neighbor relationships, resulting in the loss of local context. This pape...
Neighbor-Aware Token Reduction via Hilbert Curve for Vision Transformers. Vision Transformers (ViTs) have achieved remarkable success in visual recognition tasks, but redundant token representations limit their computational efficiency. Existing token merging and pruning strategies often overlook spatial continuity and...
cv_vision
cs.CV
2025-12-28T03:25:45+00:00
2509.08216v1
Vector embedding of multi-modal texts: a tool for discovery?
Computer science texts are particularly rich in both narrative content and illustrative charts, algorithms, images, annotated diagrams, etc. This study explores the extent to which vector-based multimodal retrieval, powered by vision-language models (VLMs), can improve discovery across multi-modal (text and images) con...
Vector embedding of multi-modal texts: a tool for discovery?. Computer science texts are particularly rich in both narrative content and illustrative charts, algorithms, images, annotated diagrams, etc. This study explores the extent to which vector-based multimodal retrieval, powered by vision-language models (VLMs), ...
ir_retrieval
cs.IR
2025-09-10T01:14:48+00:00
2510.15422v1
Information Theory in Open-world Machine Learning Foundations, Frameworks, and Future Direction
Open world Machine Learning (OWML) aims to develop intelligent systems capable of recognizing known categories, rejecting unknown samples, and continually learning from novel information. Despite significant progress in open set recognition, novelty detection, and continual learning, the field still lacks a unified the...
Information Theory in Open-world Machine Learning Foundations, Frameworks, and Future Direction. Open world Machine Learning (OWML) aims to develop intelligent systems capable of recognizing known categories, rejecting unknown samples, and continually learning from novel information. Despite significant progress in ope...
stat_ml
stat.ML
2025-10-17T08:20:56+00:00
2512.18496v1
Adaptive-VoCo: Complexity-Aware Visual Token Compression for Vision-Language Models
In recent years, large-scale vision-language models (VLMs) have demonstrated remarkable performance on multimodal understanding and reasoning tasks. However, handling high-dimensional visual features often incurs substantial computational and memory costs. VoCo-LLaMA alleviates this issue by compressing visual patch to...
Adaptive-VoCo: Complexity-Aware Visual Token Compression for Vision-Language Models. In recent years, large-scale vision-language models (VLMs) have demonstrated remarkable performance on multimodal understanding and reasoning tasks. However, handling high-dimensional visual features often incurs substantial computatio...
cv_vision
cs.CV
2025-12-20T20:24:07+00:00
2511.01404v1
A Soft-partitioned Semi-supervised Collaborative Transfer Learning Approach for Multi-Domain Recommendation
In industrial practice, Multi-domain Recommendation (MDR) plays a crucial role. Shared-specific architectures are widely used in industrial solutions to capture shared and unique attributes via shared and specific parameters. However, with imbalanced data across different domains, these models face two key issues: (1) ...
A Soft-partitioned Semi-supervised Collaborative Transfer Learning Approach for Multi-Domain Recommendation. In industrial practice, Multi-domain Recommendation (MDR) plays a crucial role. Shared-specific architectures are widely used in industrial solutions to capture shared and unique attributes via shared and specif...
ir_retrieval
cs.IR
2025-11-03T09:58:32+00:00
2512.23260v2
Interpretable Safety Alignment via SAE-Constructed Low-Rank Subspace Adaptation
Safety alignment -- training large language models (LLMs) to refuse harmful requests while remaining helpful -- is critical for responsible deployment. Prior work established that safety behaviors are governed by low-rank structures, suggesting parameter-efficient fine-tuning (PEFT) should be well-suited for alignment....
Interpretable Safety Alignment via SAE-Constructed Low-Rank Subspace Adaptation. Safety alignment -- training large language models (LLMs) to refuse harmful requests while remaining helpful -- is critical for responsible deployment. Prior work established that safety behaviors are governed by low-rank structures, sugge...
cl_nlp
cs.CL
2025-12-29T07:39:49+00:00
2512.22303v1
Attack-Aware Deepfake Detection under Counter-Forensic Manipulations
This work presents an attack-aware deepfake and image-forensics detector designed for robustness, well-calibrated probabilities, and transparent evidence under realistic deployment conditions. The method combines red-team training with randomized test-time defense in a two-stream architecture, where one stream encodes ...
Attack-Aware Deepfake Detection under Counter-Forensic Manipulations. This work presents an attack-aware deepfake and image-forensics detector designed for robustness, well-calibrated probabilities, and transparent evidence under realistic deployment conditions. The method combines red-team training with randomized tes...
cv_vision
cs.CV
2025-12-26T04:05:52+00:00
2510.14413v2
Row-wise Fusion Regularization: An Interpretable Personalized Federated Learning Framework in Large-Scale Scenarios
We study personalized federated learning for multivariate responses where client models are heterogeneous yet share variable-level structure. Existing entry-wise penalties ignore cross-response dependence, while matrix-wise fusion over-couples clients. We propose a Sparse Row-wise Fusion (SROF) regularizer that cluster...
Row-wise Fusion Regularization: An Interpretable Personalized Federated Learning Framework in Large-Scale Scenarios. We study personalized federated learning for multivariate responses where client models are heterogeneous yet share variable-level structure. Existing entry-wise penalties ignore cross-response dependenc...
stat_ml
stat.ML
2025-10-16T08:18:36+00:00
2512.17312v2
CodeDance: A Dynamic Tool-integrated MLLM for Executable Visual Reasoning
Recent releases such as o3 highlight human-like "thinking with images" reasoning that combines tool use with stepwise verification, yet most open-source approaches still rely on text-only chains, rigid visual schemas, or single-step pipelines, limiting flexibility, interpretability, and transferability on complex tasks...
CodeDance: A Dynamic Tool-integrated MLLM for Executable Visual Reasoning. Recent releases such as o3 highlight human-like "thinking with images" reasoning that combines tool use with stepwise verification, yet most open-source approaches still rely on text-only chains, rigid visual schemas, or single-step pipelines, l...
cv_vision
cs.CV
2025-12-19T07:52:23+00:00
2511.13911v1
Uncertainty-Calibrated Prediction of Randomly-Timed Biomarker Trajectories with Conformal Bands
Despite recent progress in predicting biomarker trajectories from real clinical data, uncertainty in the predictions poses high-stakes risks (e.g., misdiagnosis) that limit their clinical deployment. To enable safe and reliable use of such predictions in healthcare, we introduce a conformal method for uncertainty-calib...
Uncertainty-Calibrated Prediction of Randomly-Timed Biomarker Trajectories with Conformal Bands. Despite recent progress in predicting biomarker trajectories from real clinical data, uncertainty in the predictions poses high-stakes risks (e.g., misdiagnosis) that limit their clinical deployment. To enable safe and reli...
stat_ml
stat.ML
2025-11-17T21:04:14+00:00
2511.10138v3
GPR: Towards a Generative Pre-trained One-Model Paradigm for Large-Scale Advertising Recommendation
As an intelligent infrastructure connecting users with commercial content, advertising recommendation systems play a central role in information flow and value creation within the digital economy. However, existing multi-stage advertising recommendation systems suffer from objective misalignment and error propagation, ...
GPR: Towards a Generative Pre-trained One-Model Paradigm for Large-Scale Advertising Recommendation. As an intelligent infrastructure connecting users with commercial content, advertising recommendation systems play a central role in information flow and value creation within the digital economy. However, existing mult...
ir_retrieval
cs.IR
2025-11-13T09:50:53+00:00
2512.17985v1
MoE-TransMov: A Transformer-based Model for Next POI Prediction in Familiar & Unfamiliar Movements
Accurate prediction of the next point of interest (POI) within human mobility trajectories is essential for location-based services, as it enables more timely and personalized recommendations. In particular, with the rise of these approaches, studies have shown that users exhibit different POI choices in their familiar...
MoE-TransMov: A Transformer-based Model for Next POI Prediction in Familiar & Unfamiliar Movements. Accurate prediction of the next point of interest (POI) within human mobility trajectories is essential for location-based services, as it enables more timely and personalized recommendations. In particular, with the ris...
lg_ml
cs.LG
2025-12-19T15:03:49+00:00
2512.18003v2
Name That Part: 3D Part Segmentation and Naming
We address semantic 3D part segmentation: decomposing objects into parts with meaningful names. While datasets exist with part annotations, their definitions are inconsistent across datasets, limiting robust training. Previous methods produce unlabeled decompositions or retrieve single parts without complete shape anno...
Name That Part: 3D Part Segmentation and Naming. We address semantic 3D part segmentation: decomposing objects into parts with meaningful names. While datasets exist with part annotations, their definitions are inconsistent across datasets, limiting robust training. Previous methods produce unlabeled decompositions or ...
cv_vision
cs.CV
2025-12-19T19:02:36+00:00
2512.09149v2
MindShift: Analyzing Language Models' Reactions to Psychological Prompts
Large language models (LLMs) hold the potential to absorb and reflect personality traits and attitudes specified by users. In our study, we investigated this potential using robust psychometric measures. We adapted the most studied test in psychological literature, namely Minnesota Multiphasic Personality Inventory (MM...
MindShift: Analyzing Language Models' Reactions to Psychological Prompts. Large language models (LLMs) hold the potential to absorb and reflect personality traits and attitudes specified by users. In our study, we investigated this potential using robust psychometric measures. We adapted the most studied test in psycho...
cl_nlp
cs.CL
2025-12-09T21:56:54+00:00
2510.15814v1
On Universality of Deep Equivariant Networks
Universality results for equivariant neural networks remain rare. Those that do exist typically hold only in restrictive settings: either they rely on regular or higher-order tensor representations, leading to impractically high-dimensional hidden spaces, or they target specialized architectures, often confined to the ...
On Universality of Deep Equivariant Networks. Universality results for equivariant neural networks remain rare. Those that do exist typically hold only in restrictive settings: either they rely on regular or higher-order tensor representations, leading to impractically high-dimensional hidden spaces, or they target spe...
stat_ml
stat.ML
2025-10-17T16:51:31+00:00
2512.20821v3
Divided We Fall: Defending Against Adversarial Attacks via Soft-Gated Fractional Mixture-of-Experts with Randomized Adversarial Training
Machine learning is a powerful tool enabling full automation of a huge number of tasks without explicit programming. Despite recent progress of machine learning in different domains, these models have shown vulnerabilities when they are exposed to adversarial threats. Adversarial threats aim to hinder the machine learn...
Divided We Fall: Defending Against Adversarial Attacks via Soft-Gated Fractional Mixture-of-Experts with Randomized Adversarial Training. Machine learning is a powerful tool enabling full automation of a huge number of tasks without explicit programming. Despite recent progress of machine learning in different domains,...
lg_ml
cs.LG
2025-12-23T22:46:06+00:00
2512.07761v3
TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards
Large language models have seen widespread adoption, yet they remain vulnerable to multi-turn jailbreak attacks, threatening their safe deployment. This has led to the task of training automated multi-turn attackers to probe model safety vulnerabilities. However, existing approaches typically rely on turn-level optimiz...
TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards. Large language models have seen widespread adoption, yet they remain vulnerable to multi-turn jailbreak attacks, threatening their safe deployment. This has led to the task of training automated multi-turn attack...
ai_general
cs.AI
2025-12-08T17:42:59+00:00
2512.21094v1
T2AV-Compass: Towards Unified Evaluation for Text-to-Audio-Video Generation
Text-to-Audio-Video (T2AV) generation aims to synthesize temporally coherent video and semantically synchronized audio from natural language, yet its evaluation remains fragmented, often relying on unimodal metrics or narrowly scoped benchmarks that fail to capture cross-modal alignment, instruction following, and perc...
T2AV-Compass: Towards Unified Evaluation for Text-to-Audio-Video Generation. Text-to-Audio-Video (T2AV) generation aims to synthesize temporally coherent video and semantically synchronized audio from natural language, yet its evaluation remains fragmented, often relying on unimodal metrics or narrowly scoped benchmark...
cv_vision
cs.CV
2025-12-24T10:30:35+00:00
2512.20251v2
Degradation-Aware Metric Prompting for Hyperspectral Image Restoration
Unified hyperspectral image (HSI) restoration aims to recover various degraded HSIs using a single model, offering great practical value. However, existing methods often depend on explicit degradation priors (e.g., degradation labels) as prompts to guide restoration, which are difficult to obtain due to complex and mix...
Degradation-Aware Metric Prompting for Hyperspectral Image Restoration. Unified hyperspectral image (HSI) restoration aims to recover various degraded HSIs using a single model, offering great practical value. However, existing methods often depend on explicit degradation priors (e.g., degradation labels) as prompts to...
cv_vision
cs.CV
2025-12-23T11:05:36+00:00
2512.13762v1
State-Dependent Refusal and Learned Incapacity in RLHF-Aligned Language Models
Large language models (LLMs) are widely deployed as general-purpose tools, yet extended interaction can reveal behavioral patterns not captured by standard quantitative benchmarks. We present a qualitative case-study methodology for auditing policy-linked behavioral selectivity in long-horizon interaction. In a single ...
State-Dependent Refusal and Learned Incapacity in RLHF-Aligned Language Models. Large language models (LLMs) are widely deployed as general-purpose tools, yet extended interaction can reveal behavioral patterns not captured by standard quantitative benchmarks. We present a qualitative case-study methodology for auditin...
ai_general
cs.AI
2025-12-15T14:00:15+00:00
2512.22317v3
LangPrecip: Language-Aware Multimodal Precipitation Nowcasting
Short-term precipitation nowcasting is an inherently uncertain and under-constrained spatiotemporal forecasting problem, especially for rapidly evolving and extreme weather events. Existing generative approaches rely primarily on visual conditioning, leaving future motion weakly constrained and ambiguous. We propose a ...
LangPrecip: Language-Aware Multimodal Precipitation Nowcasting. Short-term precipitation nowcasting is an inherently uncertain and under-constrained spatiotemporal forecasting problem, especially for rapidly evolving and extreme weather events. Existing generative approaches rely primarily on visual conditioning, leavi...
lg_ml
cs.LG
2025-12-26T12:06:29+00:00
2509.19057v2
RELATE: Relation Extraction in Biomedical Abstracts with LLMs and Ontology Constraints
Biomedical knowledge graphs (KGs) are vital for drug discovery and clinical decision support but remain incomplete. Large language models (LLMs) excel at extracting biomedical relations, yet their outputs lack standardization and alignment with ontologies, limiting KG integration. We introduce RELATE, a three-stage pip...
RELATE: Relation Extraction in Biomedical Abstracts with LLMs and Ontology Constraints. Biomedical knowledge graphs (KGs) are vital for drug discovery and clinical decision support but remain incomplete. Large language models (LLMs) excel at extracting biomedical relations, yet their outputs lack standardization and al...
ir_retrieval
cs.IR
2025-09-23T14:21:46+00:00
2509.04919v1
Optimal Variance and Covariance Estimation under Differential Privacy in the Add-Remove Model and Beyond
In this paper, we study the problem of estimating the variance and covariance of datasets under differential privacy in the add-remove model. While estimation in the swap model has been extensively studied in the literature, the add-remove model remains less explored and more challenging, as the dataset size must also ...
Optimal Variance and Covariance Estimation under Differential Privacy in the Add-Remove Model and Beyond. In this paper, we study the problem of estimating the variance and covariance of datasets under differential privacy in the add-remove model. While estimation in the swap model has been extensively studied in the l...
stat_ml
stat.ML
2025-09-05T08:37:30+00:00
2512.19213v1
InvCoSS: Inversion-driven Continual Self-supervised Learning in Medical Multi-modal Image Pre-training
Continual self-supervised learning (CSSL) in medical imaging trains a foundation model sequentially, alleviating the need for collecting multi-modal images for joint training and offering promising improvements in downstream performance while preserving data privacy. However, most existing methods still rely on replayi...
InvCoSS: Inversion-driven Continual Self-supervised Learning in Medical Multi-modal Image Pre-training. Continual self-supervised learning (CSSL) in medical imaging trains a foundation model sequentially, alleviating the need for collecting multi-modal images for joint training and offering promising improvements in do...
cv_vision
cs.CV
2025-12-22T09:53:38+00:00
2512.06196v1
ARCANE: A Multi-Agent Framework for Interpretable and Configurable Alignment
As agents based on large language models are increasingly deployed to long-horizon tasks, maintaining their alignment with stakeholder preferences becomes critical. Effective alignment in such settings requires reward models that are interpretable so that stakeholders can understand and audit model objectives. Moreover...
ARCANE: A Multi-Agent Framework for Interpretable and Configurable Alignment. As agents based on large language models are increasingly deployed to long-horizon tasks, maintaining their alignment with stakeholder preferences becomes critical. Effective alignment in such settings requires reward models that are interpre...
ai_general
cs.AI
2025-12-05T22:39:54+00:00
2512.23239v2
RS-Prune: Training-Free Data Pruning at High Ratios for Efficient Remote Sensing Diffusion Foundation Models
Diffusion-based remote sensing (RS) generative foundation models are cruial for downstream tasks. However, these models rely on large amounts of globally representative data, which often contain redundancy, noise, and class imbalance, reducing training efficiency and preventing convergence. Existing RS diffusion founda...
RS-Prune: Training-Free Data Pruning at High Ratios for Efficient Remote Sensing Diffusion Foundation Models. Diffusion-based remote sensing (RS) generative foundation models are cruial for downstream tasks. However, these models rely on large amounts of globally representative data, which often contain redundancy, noi...
cv_vision
cs.CV
2025-12-29T06:44:06+00:00
2601.00894v1
When to Ponder: Adaptive Compute Allocation for Code Generation via Test-Time Training
Large language models apply uniform computation to all inputs, regardless of difficulty. We propose PonderTTT, a gating strategy using the TTT layer's self-supervised reconstruction loss to selectively trigger Test-Time Training (TTT) updates. The gating decision itself is training-free--requiring no learned classifier...
When to Ponder: Adaptive Compute Allocation for Code Generation via Test-Time Training. Large language models apply uniform computation to all inputs, regardless of difficulty. We propose PonderTTT, a gating strategy using the TTT layer's self-supervised reconstruction loss to selectively trigger Test-Time Training (TT...
lg_ml
cs.LG
2025-12-31T14:49:54+00:00
2510.24431v1
MiniOneRec: An Open-Source Framework for Scaling Generative Recommendation
The recent success of large language models (LLMs) has renewed interest in whether recommender systems can achieve similar scaling benefits. Conventional recommenders, dominated by massive embedding tables, tend to plateau as embedding dimensions grow. In contrast, the emerging generative paradigm replaces embeddings w...
MiniOneRec: An Open-Source Framework for Scaling Generative Recommendation. The recent success of large language models (LLMs) has renewed interest in whether recommender systems can achieve similar scaling benefits. Conventional recommenders, dominated by massive embedding tables, tend to plateau as embedding dimensio...
ir_retrieval
cs.IR
2025-10-28T13:58:36+00:00
2512.16248v2
Sigma-MoE-Tiny Technical Report
Mixture-of-Experts (MoE) has emerged as a promising paradigm for foundation models due to its efficient and powerful scalability. In this work, we present Sigma-MoE-Tiny, an MoE language model that achieves the highest sparsity compared to existing open-source models. Sigma-MoE-Tiny employs fine-grained expert segmenta...
Sigma-MoE-Tiny Technical Report. Mixture-of-Experts (MoE) has emerged as a promising paradigm for foundation models due to its efficient and powerful scalability. In this work, we present Sigma-MoE-Tiny, an MoE language model that achieves the highest sparsity compared to existing open-source models. Sigma-MoE-Tiny emp...
cl_nlp
cs.CL
2025-12-18T06:57:42+00:00
2512.08366v2
Reflecting with Two Voices: A Co-Adaptive Dual-Strategy Framework for LLM-Based Agent Decision Making
Large language model (LLM) agents often rely on external demonstrations or retrieval-augmented planning, leading to brittleness, poor generalization, and high computational overhead. Inspired by human problem-solving, we propose DuSAR (Dual-Strategy Agent with Reflecting) -- a demonstration-free framework that enables ...
Reflecting with Two Voices: A Co-Adaptive Dual-Strategy Framework for LLM-Based Agent Decision Making. Large language model (LLM) agents often rely on external demonstrations or retrieval-augmented planning, leading to brittleness, poor generalization, and high computational overhead. Inspired by human problem-solving,...
ai_general
cs.AI
2025-12-09T08:44:59+00:00
2512.23343v1
AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agents
Memory serves as the pivotal nexus bridging past and future, providing both humans and AI systems with invaluable concepts and experience to navigate complex tasks. Recent research on autonomous agents has increasingly focused on designing efficient memory workflows by drawing on cognitive neuroscience. However, constr...
AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agents. Memory serves as the pivotal nexus bridging past and future, providing both humans and AI systems with invaluable concepts and experience to navigate complex tasks. Recent research on autonomous agents has increasingly focused on designing...
cl_nlp
cs.CL
2025-12-29T10:01:32+00:00
2512.23292v3
Agentic Physical AI toward a Domain-Specific Foundation Model for Nuclear Reactor Control
The prevailing paradigm in AI for physical systems (scaling general-purpose foundation models toward universal multimodal reasoning) confronts a fundamental barrier at the control interface. Recent benchmarks show that even frontier vision--language models achieve only 50--53% accuracy on basic quantitative physics tas...
Agentic Physical AI toward a Domain-Specific Foundation Model for Nuclear Reactor Control. The prevailing paradigm in AI for physical systems (scaling general-purpose foundation models toward universal multimodal reasoning) confronts a fundamental barrier at the control interface. Recent benchmarks show that even front...
ai_general
cs.AI
2025-12-29T08:26:27+00:00
2512.20582v1
Relu and softplus neural nets as zero-sum turn-based games
We show that the output of a ReLU neural network can be interpreted as the value of a zero-sum, turn-based, stopping game, which we call the ReLU net game. The game runs in the direction opposite to that of the network, and the input of the network serves as the terminal reward of the game. In fact, evaluating the netw...
Relu and softplus neural nets as zero-sum turn-based games. We show that the output of a ReLU neural network can be interpreted as the value of a zero-sum, turn-based, stopping game, which we call the ReLU net game. The game runs in the direction opposite to that of the network, and the input of the network serves as t...
lg_ml
cs.LG
2025-12-23T18:27:41+00:00
2512.22106v1
Pruning as a Game: Equilibrium-Driven Sparsification of Neural Networks
Neural network pruning is widely used to reduce model size and computational cost. Yet, most existing methods treat sparsity as an externally imposed constraint, enforced through heuristic importance scores or training-time regularization. In this work, we propose a fundamentally different perspective: pruning as an eq...
Pruning as a Game: Equilibrium-Driven Sparsification of Neural Networks. Neural network pruning is widely used to reduce model size and computational cost. Yet, most existing methods treat sparsity as an externally imposed constraint, enforced through heuristic importance scores or training-time regularization. In this...
ai_general
cs.AI
2025-12-26T18:25:38+00:00
2512.10963v1
Emotion-Driven Personalized Recommendation for AI-Generated Content Using Multi-Modal Sentiment and Intent Analysis
With the rapid growth of AI-generated content (AIGC) across domains such as music, video, and literature, the demand for emotionally aware recommendation systems has become increasingly important. Traditional recommender systems primarily rely on user behavioral data such as clicks, views, or ratings, while neglecting ...
Emotion-Driven Personalized Recommendation for AI-Generated Content Using Multi-Modal Sentiment and Intent Analysis. With the rapid growth of AI-generated content (AIGC) across domains such as music, video, and literature, the demand for emotionally aware recommendation systems has become increasingly important. Tradit...
ir_retrieval
cs.IR
2025-11-25T17:52:22+00:00
2510.18259v3
Learning under Quantization for High-Dimensional Linear Regression
The use of low-bit quantization has emerged as an indispensable technique for enabling the efficient training of large-scale models. Despite its widespread empirical success, a rigorous theoretical understanding of its impact on learning performance remains notably absent, even in the simplest linear regression setting...
Learning under Quantization for High-Dimensional Linear Regression. The use of low-bit quantization has emerged as an indispensable technique for enabling the efficient training of large-scale models. Despite its widespread empirical success, a rigorous theoretical understanding of its impact on learning performance re...
stat_ml
stat.ML
2025-10-21T03:30:11+00:00
2512.19913v1
Quasiprobabilistic Density Ratio Estimation with a Reverse Engineered Classification Loss Function
We consider a generalization of the classifier-based density-ratio estimation task to a quasiprobabilistic setting where probability densities can be negative. The problem with most loss functions used for this task is that they implicitly define a relationship between the optimal classifier and the target quasiprobabi...
Quasiprobabilistic Density Ratio Estimation with a Reverse Engineered Classification Loss Function. We consider a generalization of the classifier-based density-ratio estimation task to a quasiprobabilistic setting where probability densities can be negative. The problem with most loss functions used for this task is t...
stat_ml
stat.ML
2025-12-22T22:37:19+00:00
2512.20573v3
Fail Fast, Win Big: Rethinking the Drafting Strategy in Speculative Decoding via Diffusion LLMs
Diffusion Large Language Models (dLLMs) offer fast, parallel token generation, but their standalone use is plagued by an inherent efficiency-quality tradeoff. We show that, if carefully applied, the attributes of dLLMs can actually be a strength for drafters in speculative decoding with autoregressive (AR) verifiers. O...
Fail Fast, Win Big: Rethinking the Drafting Strategy in Speculative Decoding via Diffusion LLMs. Diffusion Large Language Models (dLLMs) offer fast, parallel token generation, but their standalone use is plagued by an inherent efficiency-quality tradeoff. We show that, if carefully applied, the attributes of dLLMs can ...
lg_ml
cs.LG
2025-12-23T18:16:58+00:00
2512.09636v2
MentraSuite: Post-Training Large Language Models for Mental Health Reasoning and Assessment
Mental health disorders affect hundreds of millions globally, and the Web now serves as a primary medium for accessing support, information, and assessment. Large language models (LLMs) offer scalable and accessible assistance, yet their deployment in mental-health settings remains risky when their reasoning is incompl...
MentraSuite: Post-Training Large Language Models for Mental Health Reasoning and Assessment. Mental health disorders affect hundreds of millions globally, and the Web now serves as a primary medium for accessing support, information, and assessment. Large language models (LLMs) offer scalable and accessible assistance,...
cl_nlp
cs.CL
2025-12-10T13:26:22+00:00
2601.02372v1
Improving News Recommendations through Hybrid Sentiment Modelling and Reinforcement Learning
News recommendation systems rely on automated sentiment analysis to personalise content and enhance user engagement. Conventional approaches often struggle with ambiguity, lexicon inconsistencies, and limited contextual understanding, particularly in multi-source news environments. Existing models typically treat senti...
Improving News Recommendations through Hybrid Sentiment Modelling and Reinforcement Learning. News recommendation systems rely on automated sentiment analysis to personalise content and enhance user engagement. Conventional approaches often struggle with ambiguity, lexicon inconsistencies, and limited contextual unders...
ir_retrieval
cs.IR
2025-12-08T21:39:11+00:00
2512.17161v1
Distributed Learning in Markovian Restless Bandits over Interference Graphs for Stable Spectrum Sharing
We study distributed learning for spectrum access and sharing among multiple cognitive communication entities, such as cells, subnetworks, or cognitive radio users (collectively referred to as cells), in communication-constrained wireless networks modeled by interference graphs. Our goal is to achieve a globally stable...
Distributed Learning in Markovian Restless Bandits over Interference Graphs for Stable Spectrum Sharing. We study distributed learning for spectrum access and sharing among multiple cognitive communication entities, such as cells, subnetworks, or cognitive radio users (collectively referred to as cells), in communicati...
lg_ml
cs.LG
2025-12-19T01:43:00+00:00
2512.24853v1
A study on constraint extraction and exception exclusion in care worker scheduling
Technologies for automatically generating work schedules have been extensively studied; however, in long-term care facilities, the conditions vary between facilities, making it essential to interview the managers who create shift schedules to design facility-specific constraint conditions. The proposed method utilizes ...
A study on constraint extraction and exception exclusion in care worker scheduling. Technologies for automatically generating work schedules have been extensively studied; however, in long-term care facilities, the conditions vary between facilities, making it essential to interview the managers who create shift schedu...
ai_general
cs.AI
2025-12-31T13:22:50+00:00
2510.05598v3
AgentDR: Dynamic Recommendation with Implicit Item-Item Relations via LLM-based Agents
Recent agent-based recommendation frameworks aim to simulate user behaviors by incorporating memory mechanisms and prompting strategies, but they struggle with hallucinating non-existent items and full-catalog ranking. Besides, a largely underexplored opportunity lies in leveraging LLMs'commonsense reasoning to capture...
AgentDR: Dynamic Recommendation with Implicit Item-Item Relations via LLM-based Agents. Recent agent-based recommendation frameworks aim to simulate user behaviors by incorporating memory mechanisms and prompting strategies, but they struggle with hallucinating non-existent items and full-catalog ranking. Besides, a la...
ir_retrieval
cs.IR
2025-10-07T05:48:05+00:00
2512.24077v1
LoongFlow: Directed Evolutionary Search via a Cognitive Plan-Execute-Summarize Paradigm
The transition from static Large Language Models (LLMs) to self-improving agents is hindered by the lack of structured reasoning in traditional evolutionary approaches. Existing methods often struggle with premature convergence and inefficient exploration in high-dimensional code spaces. To address these challenges, we...
LoongFlow: Directed Evolutionary Search via a Cognitive Plan-Execute-Summarize Paradigm. The transition from static Large Language Models (LLMs) to self-improving agents is hindered by the lack of structured reasoning in traditional evolutionary approaches. Existing methods often struggle with premature convergence and...
ai_general
cs.AI
2025-12-30T08:39:28+00:00
2512.18843v1
Brain-Gen: Towards Interpreting Neural Signals for Stimulus Reconstruction Using Transformers and Latent Diffusion Models
Advances in neuroscience and artificial intelligence have enabled preliminary decoding of brain activity. However, despite the progress, the interpretability of neural representations remains limited. A significant challenge arises from the intrinsic properties of electroencephalography (EEG) signals, including high no...
Brain-Gen: Towards Interpreting Neural Signals for Stimulus Reconstruction Using Transformers and Latent Diffusion Models. Advances in neuroscience and artificial intelligence have enabled preliminary decoding of brain activity. However, despite the progress, the interpretability of neural representations remains limit...
cv_vision
cs.CV
2025-12-21T18:20:21+00:00
2512.19534v1
SlicerOrbitSurgerySim: An Open-Source Platform for Virtual Registration and Quantitative Comparison of Preformed Orbital Plates
Poor adaptation of orbital implants remains a major contributor to postoperative complications and revision surgery. Although preformed orbital plates are widely used to reduce cost and operative time compared with customized implants, surgeons currently lack publicly available tools and standardized metrics to quantit...
SlicerOrbitSurgerySim: An Open-Source Platform for Virtual Registration and Quantitative Comparison of Preformed Orbital Plates. Poor adaptation of orbital implants remains a major contributor to postoperative complications and revision surgery. Although preformed orbital plates are widely used to reduce cost and opera...
cv_vision
cs.CV
2025-12-22T16:21:29+00:00
2512.22087v1
Context as a Tool: Context Management for Long-Horizon SWE-Agents
Agents based on large language models have recently shown strong potential on real-world software engineering (SWE) tasks that require long-horizon interaction with repository-scale codebases. However, most existing agents rely on append-only context maintenance or passively triggered compression heuristics, which ofte...
Context as a Tool: Context Management for Long-Horizon SWE-Agents. Agents based on large language models have recently shown strong potential on real-world software engineering (SWE) tasks that require long-horizon interaction with repository-scale codebases. However, most existing agents rely on append-only context ma...
cl_nlp
cs.CL
2025-12-26T17:15:47+00:00
2512.22236v1
We are not able to identify AI-generated images
AI-generated images are now pervasive online, yet many people believe they can easily tell them apart from real photographs. We test this assumption through an interactive web experiment where participants classify 20 images as real or AI-generated. Our dataset contains 120 difficult cases: real images sampled from CC1...
We are not able to identify AI-generated images. AI-generated images are now pervasive online, yet many people believe they can easily tell them apart from real photographs. We test this assumption through an interactive web experiment where participants classify 20 images as real or AI-generated. Our dataset contains ...
ai_general
cs.AI
2025-12-23T11:55:40+00:00
2512.18683v1
CIRR: Causal-Invariant Retrieval-Augmented Recommendation with Faithful Explanations under Distribution Shift
Recent advances in retrieval-augmented generation (RAG) have shown promise in enhancing recommendation systems with external knowledge. However, existing RAG-based recommenders face two critical challenges: (1) vulnerability to distribution shifts across different environments (e.g., time periods, user segments), leadi...
CIRR: Causal-Invariant Retrieval-Augmented Recommendation with Faithful Explanations under Distribution Shift. Recent advances in retrieval-augmented generation (RAG) have shown promise in enhancing recommendation systems with external knowledge. However, existing RAG-based recommenders face two critical challenges: (1...
ir_retrieval
cs.IR
2025-12-21T10:41:05+00:00
2510.20455v1
Rotate Both Ways: Time-and-Order RoPE for Generative Recommendation
Generative recommenders, typically transformer-based autoregressive models, predict the next item or action from a user's interaction history. Their effectiveness depends on how the model represents where an interaction event occurs in the sequence (discrete index) and when it occurred in wall-clock time. Prevailing ap...
Rotate Both Ways: Time-and-Order RoPE for Generative Recommendation. Generative recommenders, typically transformer-based autoregressive models, predict the next item or action from a user's interaction history. Their effectiveness depends on how the model represents where an interaction event occurs in the sequence (d...
ir_retrieval
cs.IR
2025-10-23T11:44:56+00:00
2512.20732v1
FEM-Bench: A Structured Scientific Reasoning Benchmark for Evaluating Code-Generating LLMs
As LLMs advance their reasoning capabilities about the physical world, the absence of rigorous benchmarks for evaluating their ability to generate scientifically valid physical models has become a critical gap. Computational mechanics, which develops and applies mathematical models and numerical methods to predict the ...
FEM-Bench: A Structured Scientific Reasoning Benchmark for Evaluating Code-Generating LLMs. As LLMs advance their reasoning capabilities about the physical world, the absence of rigorous benchmarks for evaluating their ability to generate scientifically valid physical models has become a critical gap. Computational mec...
lg_ml
cs.LG
2025-12-23T19:40:51+00:00
2512.12264v2
Market-Bench: Evaluating Large Language Models on Introductory Quantitative Trading and Market Dynamics
We introduce MARKET-BENCH, a benchmark that evaluates large language models (LLMs) on introductory quantitative trading tasks by asking them to construct executable backtesters from natural language strategy descriptions and market assumptions. Each instance specifies one of three canonical strategies: scheduled tradin...
Market-Bench: Evaluating Large Language Models on Introductory Quantitative Trading and Market Dynamics. We introduce MARKET-BENCH, a benchmark that evaluates large language models (LLMs) on introductory quantitative trading tasks by asking them to construct executable backtesters from natural language strategy descrip...
cl_nlp
cs.CL
2025-12-13T10:07:31+00:00
2512.24622v1
FireRescue: A UAV-Based Dataset and Enhanced YOLO Model for Object Detection in Fire Rescue Scenes
Object detection in fire rescue scenarios is importance for command and decision-making in firefighting operations. However, existing research still suffers from two main limitations. First, current work predominantly focuses on environments such as mountainous or forest areas, while paying insufficient attention to ur...
FireRescue: A UAV-Based Dataset and Enhanced YOLO Model for Object Detection in Fire Rescue Scenes. Object detection in fire rescue scenarios is importance for command and decision-making in firefighting operations. However, existing research still suffers from two main limitations. First, current work predominantly fo...
cv_vision
cs.CV
2025-12-31T04:37:51+00:00
2509.00265v2
The Nondecreasing Rank
In this article the notion of the nondecreasing (ND) rank of a matrix or tensor is introduced. A tensor has an ND rank of r if it can be represented as a sum of r outer products of vectors, with each vector satisfying a monotonicity constraint. It is shown that for certain poset orderings finding an ND factorization of...
The Nondecreasing Rank. In this article the notion of the nondecreasing (ND) rank of a matrix or tensor is introduced. A tensor has an ND rank of r if it can be represented as a sum of r outer products of vectors, with each vector satisfying a monotonicity constraint. It is shown that for certain poset orderings findin...
stat_ml
stat.ML
2025-08-29T22:31:04+00:00
2512.13565v1
A Nonparametric Statistics Approach to Feature Selection in Deep Neural Networks with Theoretical Guarantees
This paper tackles the problem of feature selection in a highly challenging setting: $\mathbb{E}(y | \boldsymbol{x}) = G(\boldsymbol{x}_{\mathcal{S}_0})$, where $\mathcal{S}_0$ is the set of relevant features and $G$ is an unknown, potentially nonlinear function subject to mild smoothness conditions. Our approach begin...
A Nonparametric Statistics Approach to Feature Selection in Deep Neural Networks with Theoretical Guarantees. This paper tackles the problem of feature selection in a highly challenging setting: $\mathbb{E}(y | \boldsymbol{x}) = G(\boldsymbol{x}_{\mathcal{S}_0})$, where $\mathcal{S}_0$ is the set of relevant features a...
stat_ml
stat.ML
2025-12-15T17:22:49+00:00
2512.22895v1
SAMP-HDRL: Segmented Allocation with Momentum-Adjusted Utility for Multi-agent Portfolio Management via Hierarchical Deep Reinforcement Learning
Portfolio optimization in non-stationary markets is challenging due to regime shifts, dynamic correlations, and the limited interpretability of deep reinforcement learning (DRL) policies. We propose a Segmented Allocation with Momentum-Adjusted Utility for Multi-agent Portfolio Management via Hierarchical Deep Reinforc...
SAMP-HDRL: Segmented Allocation with Momentum-Adjusted Utility for Multi-agent Portfolio Management via Hierarchical Deep Reinforcement Learning. Portfolio optimization in non-stationary markets is challenging due to regime shifts, dynamic correlations, and the limited interpretability of deep reinforcement learning (D...
ai_general
cs.AI
2025-12-28T11:56:39+00:00
2601.00885v1
Counterfactual Self-Questioning for Stable Policy Optimization in Language Models
Recent work on language model self-improvement shows that models can refine their own reasoning through reflection, verification, debate, or self-generated rewards. However, most existing approaches rely on external critics, learned reward models, or ensemble sampling, which increases complexity and training instabilit...
Counterfactual Self-Questioning for Stable Policy Optimization in Language Models. Recent work on language model self-improvement shows that models can refine their own reasoning through reflection, verification, debate, or self-generated rewards. However, most existing approaches rely on external critics, learned rewa...
ai_general
cs.AI
2025-12-31T09:10:37+00:00
2510.16728v1
Local regression on path spaces with signature metrics
We study nonparametric regression and classification for path-valued data. We introduce a functional Nadaraya-Watson estimator that combines the signature transform from rough path theory with local kernel regression. The signature transform provides a principled way to encode sequential data through iterated integrals...
Local regression on path spaces with signature metrics. We study nonparametric regression and classification for path-valued data. We introduce a functional Nadaraya-Watson estimator that combines the signature transform from rough path theory with local kernel regression. The signature transform provides a principled ...
stat_ml
stat.ML
2025-10-19T06:40:51+00:00
2512.03293v1
Prior preferences in active inference agents: soft, hard, and goal shaping
Active inference proposes expected free energy as an objective for planning and decision-making to adequately balance exploitative and explorative drives in learning agents. The exploitative drive, or what an agent wants to achieve, is formalised as the Kullback-Leibler divergence between a variational probability dist...
Prior preferences in active inference agents: soft, hard, and goal shaping. Active inference proposes expected free energy as an objective for planning and decision-making to adequately balance exploitative and explorative drives in learning agents. The exploitative drive, or what an agent wants to achieve, is formalis...
ai_general
cs.AI
2025-12-02T23:07:24+00:00
2602.23368v1
Keyword search is all you need: Achieving RAG-Level Performance without vector databases using agentic tool use
While Retrieval-Augmented Generation (RAG) has proven effective for generating accurate, context-based responses based on existing knowledge bases, it presents several challenges including retrieval quality dependencies, integration complexity and cost. Recent advances in agentic-RAG and tool-augmented LLM architecture...
Keyword search is all you need: Achieving RAG-Level Performance without vector databases using agentic tool use. While Retrieval-Augmented Generation (RAG) has proven effective for generating accurate, context-based responses based on existing knowledge bases, it presents several challenges including retrieval quality ...
ir_retrieval
cs.IR
2025-12-19T20:15:30+00:00
2512.09895v1
Human-in-the-Loop and AI: Crowdsourcing Metadata Vocabulary for Materials Science
Metadata vocabularies are essential for advancing FAIR and FARR data principles, but their development constrained by limited human resources and inconsistent standardization practices. This paper introduces MatSci-YAMZ, a platform that integrates artificial intelligence (AI) and human-in-the-loop (HILT), including cro...
Human-in-the-Loop and AI: Crowdsourcing Metadata Vocabulary for Materials Science. Metadata vocabularies are essential for advancing FAIR and FARR data principles, but their development constrained by limited human resources and inconsistent standardization practices. This paper introduces MatSci-YAMZ, a platform that ...
ai_general
cs.AI
2025-12-10T18:22:57+00:00
2509.10245v1
Model-agnostic post-hoc explainability for recommender systems
Recommender systems often benefit from complex feature embeddings and deep learning algorithms, which deliver sophisticated recommendations that enhance user experience, engagement, and revenue. However, these methods frequently reduce the interpretability and transparency of the system. In this research, we develop a ...
Model-agnostic post-hoc explainability for recommender systems. Recommender systems often benefit from complex feature embeddings and deep learning algorithms, which deliver sophisticated recommendations that enhance user experience, engagement, and revenue. However, these methods frequently reduce the interpretability...
ir_retrieval
cs.IR
2025-09-12T13:43:16+00:00
2512.24561v1
RGBT-Ground Benchmark: Visual Grounding Beyond RGB in Complex Real-World Scenarios
Visual Grounding (VG) aims to localize specific objects in an image according to natural language expressions, serving as a fundamental task in vision-language understanding. However, existing VG benchmarks are mostly derived from datasets collected under clean environments, such as COCO, where scene diversity is limit...
RGBT-Ground Benchmark: Visual Grounding Beyond RGB in Complex Real-World Scenarios. Visual Grounding (VG) aims to localize specific objects in an image according to natural language expressions, serving as a fundamental task in vision-language understanding. However, existing VG benchmarks are mostly derived from datas...
cv_vision
cs.CV
2025-12-31T02:01:02+00:00
2512.22774v1
Schrodinger AI: A Unified Spectral-Dynamical Framework for Classification, Reasoning, and Operator-Based Generalization
We introduce \textbf{Schrödinger AI}, a unified machine learning framework inspired by quantum mechanics. The system is defined by three tightly coupled components: (1) a {time-independent wave-energy solver} that treats perception and classification as spectral decomposition under a learned Hamiltonian; (2) a {time-de...
Schrodinger AI: A Unified Spectral-Dynamical Framework for Classification, Reasoning, and Operator-Based Generalization. We introduce \textbf{Schrödinger AI}, a unified machine learning framework inspired by quantum mechanics. The system is defined by three tightly coupled components: (1) a {time-independent wave-energ...
lg_ml
cs.LG
2025-12-28T04:33:18+00:00
2512.24898v1
PRISM: A hierarchical multiscale approach for time series forecasting
Forecasting is critical in areas such as finance, biology, and healthcare. Despite the progress in the field, making accurate forecasts remains challenging because real-world time series contain both global trends, local fine-grained structure, and features on multiple scales in between. Here, we present a new forecast...
PRISM: A hierarchical multiscale approach for time series forecasting. Forecasting is critical in areas such as finance, biology, and healthcare. Despite the progress in the field, making accurate forecasts remains challenging because real-world time series contain both global trends, local fine-grained structure, and ...
lg_ml
cs.LG
2025-12-31T14:51:12+00:00
2512.24731v1
EchoFoley: Event-Centric Hierarchical Control for Video Grounded Creative Sound Generation
Sound effects build an essential layer of multimodal storytelling, shaping the emotional atmosphere and the narrative semantics of videos. Despite recent advancement in video-text-to-audio (VT2A), the current formulation faces three key limitations: First, an imbalance between visual and textual conditioning that leads...
EchoFoley: Event-Centric Hierarchical Control for Video Grounded Creative Sound Generation. Sound effects build an essential layer of multimodal storytelling, shaping the emotional atmosphere and the narrative semantics of videos. Despite recent advancement in video-text-to-audio (VT2A), the current formulation faces t...
cv_vision
cs.CV
2025-12-31T08:58:30+00:00
2512.24408v2
DyStream: Streaming Dyadic Talking Heads Generation via Flow Matching-based Autoregressive Model
Generating realistic, dyadic talking head video requires ultra-low latency. Existing chunk-based methods require full non-causal context windows, introducing significant delays. This high latency critically prevents the immediate, non-verbal feedback required for a realistic listener. To address this, we present DyStre...
DyStream: Streaming Dyadic Talking Heads Generation via Flow Matching-based Autoregressive Model. Generating realistic, dyadic talking head video requires ultra-low latency. Existing chunk-based methods require full non-causal context windows, introducing significant delays. This high latency critically prevents the im...
cv_vision
cs.CV
2025-12-30T18:43:38+00:00
2512.07218v1
NeSTR: A Neuro-Symbolic Abductive Framework for Temporal Reasoning in Large Language Models
Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, temporal reasoning, particularly under complex temporal constraints, remains a major challenge. To this end, existing approaches have explored symbolic methods, which encode temporal ...
NeSTR: A Neuro-Symbolic Abductive Framework for Temporal Reasoning in Large Language Models. Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, temporal reasoning, particularly under complex temporal constraints, remains a major chall...
cl_nlp
cs.CL
2025-12-08T06:58:23+00:00
2510.27056v1
Overspecified Mixture Discriminant Analysis: Exponential Convergence, Statistical Guarantees, and Remote Sensing Applications
This study explores the classification error of Mixture Discriminant Analysis (MDA) in scenarios where the number of mixture components exceeds those present in the actual data distribution, a condition known as overspecification. We use a two-component Gaussian mixture model within each class to fit data generated fro...
Overspecified Mixture Discriminant Analysis: Exponential Convergence, Statistical Guarantees, and Remote Sensing Applications. This study explores the classification error of Mixture Discriminant Analysis (MDA) in scenarios where the number of mixture components exceeds those present in the actual data distribution, a ...
stat_ml
stat.ML
2025-10-30T23:56:56+00:00