3/25/2026
Sparse but Critical: A Token-Level Analysis of Distributional Shifts in RLVR Fine-Tuning of LLMs
Token-level analysis of distributional shifts during RLVR fine-tuning of LLMs, examining mechanisms underlying reasoning improvements.
3/25/2026
Functional Component Ablation Reveals Specialization Patterns in Hybrid Language Model Architectures
Functional component ablation framework analyzing specialization in hybrid language models combining attention with state space models or linear attention.
3/25/2026
OrgForge-IT: A Verifiable Synthetic Benchmark for LLM-Based Insider Threat Detection
Verifiable synthetic benchmark for LLM-based insider threat detection using deterministic simulation engine to maintain ground truth and cross-artifact consistency.
3/25/2026
Privacy-Preserving Reinforcement Learning from Human Feedback via Decoupled Reward Modeling
Differential privacy framework for RLHF fine-tuning that decouples reward learning to preserve user privacy in LLM preference-based training.
3/25/2026
Benchmarking Multi-Agent LLM Architectures for Financial Document Processing: A Comparative Study of Orchestration Patterns, Cost-Accuracy Tradeoffs and Production Scaling Strategies
Systematic benchmark comparing four multi-agent LLM orchestration architectures for financial document processing with cost-accuracy tradeoffs and scaling strategies.
3/25/2026
Improving LLM Predictions via Inter-Layer Structural Encoders
Method leveraging intermediate layer representations in LLMs via Inter-Layer Structural Encoders to improve task-specific predictions beyond final-layer features.
3/25/2026
REALITrees: Rashomon Ensemble Active Learning for Interpretable Trees
Active learning approach using Rashomon ensemble for interpretable decision tree induction with direct hypothesis space characterization.
3/25/2026
KALAVAI: Predicting When Independent Specialist Fusion Works -- A Quantitative Model for Post-Hoc Cooperative LLM Training
Quantitative model predicting when independently fine-tuned specialist LLMs can be fused post-hoc for improved performance using divergence metric.
3/25/2026
Reconstruction-Guided Slot Curriculum: Addressing Object Over-Fragmentation in Video Object-Centric Learning
Method addressing over-fragmentation in video object-centric learning through reconstruction-guided slot curriculum training approach.
3/25/2026
When AI Shows Its Work, Is It Actually Working? Step-Level Evaluation Reveals Frontier Language Models Frequently Bypass Their Own Reasoning
Research on whether LLMs' step-by-step reasoning is genuinely used or post-hoc narrative generation through step-level evaluation of frontier models.
3/25/2026
TorR: Towards Brain-Inspired Task-Oriented Reasoning via Cache-Oriented Algorithm-Architecture Co-design
arXiv paper on brain-inspired object detection co-design. Algorithm-architecture optimization for CLIP-based task-oriented detection on edge devices.
3/25/2026
Dynamical Systems Theory Behind a Hierarchical Reasoning Model
arXiv paper analyzing hierarchical reasoning models for LLMs. Mathematical theory of recursive networks for algorithmic reasoning.
3/25/2026
Dual-Teacher Distillation with Subnetwork Rectification for Black-Box Domain Adaptation
arXiv paper on black-box domain adaptation using dual-teacher distillation. Technical ML research on knowledge transfer without source access.
3/25/2026
A Critical Review on the Effectiveness and Privacy Threats of Membership Inference Attacks
Critical review framework evaluating membership inference attacks and conditions under which they pose genuine privacy threats to ML models.
3/25/2026
Can Large Language Models Reason and Optimize Under Constraints?
Investigation of LLMs' reasoning and optimization capabilities under physical and operational constraints using Optimal Power Flow problems.
3/25/2026
YOLOv10 with Kolmogorov-Arnold networks and vision-language foundation models for interpretable object detection and trustworthy multimodal AI in computer vision perception
Object detection framework combining YOLOv10 with Kolmogorov-Arnold networks and vision-language models for interpretability.
3/25/2026
Prompt Amplification and Zero-Shot Late Fusion in Audio-Language Models for Speech Emotion Recognition
Zero-shot late fusion method combining audio-language models with specialist models for speech emotion recognition.
3/25/2026
Machine Learning Models for the Early Detection of Burnout in Software Engineering: a Systematic Literature Review
Systematic literature review of machine learning approaches for early detection of burnout in software engineers.
3/25/2026
HGNet: Scalable Foundation Model for Automated Knowledge Graph Generation from Scientific Literature
Scalable foundation model for automated knowledge graph generation from scientific literature using domain-specific optimization.
3/25/2026
Conformal Cross-Modal Active Learning
Active learning method leveraging vision-language foundation models for data-efficient visual recognition.
3/25/2026
Robust Safety Monitoring of Language Models via Activation Watermarking
Safety monitoring approach for LLMs using activation watermarking to detect adaptive adversarial attacks during inference.
3/25/2026
Decoding AI Authorship: Can LLMs Truly Mimic Human Style Across Literature and Politics?
Evaluation of LLMs' ability to mimic authorial styles of literary and political figures using zero-shot prompting.
3/25/2026
MemCollab: Cross-Agent Memory Collaboration via Contrastive Trajectory Distillation
Framework for sharing memory systems across heterogeneous LLM-based agents via contrastive trajectory distillation to improve knowledge reuse.
3/25/2026
Is AI Catching Up to Human Expression? Exploring Emotion, Personality, Authorship, and Linguistic Style in English and Arabic with Six Large Language Models
Study evaluating whether six LLMs can emulate emotional expression and personality traits across English and Arabic languages.
3/25/2026
Not All Tokens Are Created Equal: Query-Efficient Jailbreak Fuzzing for LLMs
Research on query-efficient jailbreak fuzzing for LLMs that identifies token importance during prompt mutation to reduce redundant searching under query constraints.
3/25/2026
Drop-In Perceptual Optimization for 3D Gaussian Splatting
Perceptual optimization strategies for 3D Gaussian Splatting using distortion losses validated via large-scale human evaluation.
3/25/2026
Contextual Graph Matching with Correlated Gaussian Features
Information-theoretic analysis of contextual graph matching with correlated Gaussian features deriving recovery thresholds.
3/25/2026
ARGENT: Adaptive Hierarchical Image-Text Representations
ARGENT: vision-language model using hyperbolic geometry to capture hierarchical structure of visual and linguistic concepts.
3/25/2026
Contrastive Metric Learning for Point Cloud Segmentation in Highly Granular Detectors
Supervised contrastive metric learning for point cloud segmentation in particle detectors using density-based clustering.
3/25/2026
VTAM: Video-Tactile-Action Models for Complex Physical Interaction Beyond VLAs
VTAM: video-tactile-action model for embodied AI combining visual and tactile signals for contact-rich physical interaction.
3/25/2026
VISion On Request: Enhanced VLLM efficiency with sparse, dynamically selected, vision-language interactions
VISOR: efficiency method for vision-language models using sparse dynamically selected interactions instead of visual token reduction.
3/25/2026
Reliable OOD Virtual Screening with Extrapolatory Pseudo-Label Matching
ML model for virtual screening in drug discovery handling out-of-distribution regions with extrapolatory pseudo-label matching.
3/25/2026
Equivariance via Minimal Frame Averaging for More Symmetries and Efficiency
Minimal Frame Averaging: framework for constructing provably minimal frames achieving exact equivariance in ML systems efficiently.
3/25/2026
Addressing Large Action Spaces in 3D Floorplanning via Spatial Generalization
Continuous action representation for 3D floorplanning addressing scalability bottlenecks from discrete canvas coordinates.
3/25/2026
Almost Sure Convergence of Linear Temporal Difference Learning with Arbitrary Features
Convergence analysis of linear temporal difference learning in reinforcement learning without requiring linearly independent features.
3/25/2026
Dataset Distillation-based Hybrid Federated Learning on Non-IID Data
HFLDD: hybrid federated learning framework using dataset distillation to address non-IID data heterogeneity and label distribution skew.
3/25/2026
DART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA
DART-Eval: benchmark for evaluating DNA language models on regulatory DNA prediction, interpretation and design tasks.
3/25/2026
MSA-CNN: A Lightweight Multi-Scale CNN with Attention for Sleep Stage Classification
MSA-CNN: lightweight multi-scale CNN with attention mechanism for automatic sleep stage classification from signal data.
3/25/2026
Streaming Attention Approximation via Discrepancy Theory
BalanceKV: streaming algorithm for approximating attention computations in LLMs using geometric process to reduce memory requirements for long-context generation.
3/25/2026
Federated Learning for Data-Driven Feedforward Control: A Case Study on Vehicle Lateral Dynamics
Federated learning approach for designing data-driven feedforward control in vehicle lateral dynamics using distributed data across multiple systems.
3/25/2026
Multiplicative learning from observation-prediction ratios
Expectation Reflection: multiplicative parameter update paradigm for ML optimization using observation-prediction ratios instead of additive gradient descent.
3/25/2026
Architecture-Aware Minimization (A$^2$M): How to Find Flat Minima in Neural Architecture Search
Analysis of geometric properties and flatness in neural architecture search spaces.
3/25/2026
Leakage and Interpretability in Concept-Based Models
Information-theoretic framework for characterizing and measuring information leakage in concept-based models.
3/25/2026
Two Stage Wireless Federated LoRA Fine-Tuning with Sparsified Orthogonal Updates
Federated LoRA fine-tuning method for large language models with communication-efficient sparsified updates.
3/25/2026
Generalizable Heuristic Generation Through LLMs with Meta-Optimization
Meta-optimization approach for generating generalizable heuristics using LLMs for combinatorial optimization.
3/25/2026
Enhancing generalizability of model discovery across parameter space with multi-experiment equation learning (ME-EQL)
Multi-experiment equation learning method for deriving analytical models from agent-based simulation data.
3/25/2026
Learning The Minimum Action Distance
State representation learning from trajectories using minimum action distance metric for MDPs.
3/25/2026
UniCA: Unified Covariate Adaptation for Time Series Foundation Model
Foundation model adapter for time series forecasting with heterogeneous covariates and multimodal data.
3/25/2026
RedTopic: Toward Topic-Diverse Red Teaming of Large Language Models
Red teaming framework for systematically discovering diverse vulnerabilities in large language models.
3/25/2026
Graph Structure Learning with Privacy Guarantees for Open Graph Data
Privacy-preserving graph structure learning using differential privacy at data publishing stage.