3/26/2026
Lagrangian Relaxation Score-based Generation for Mixed Integer linear Programming
Generative framework using Lagrangian relaxation-guided score-based generation to solve mixed-integer linear programming with diverse solutions.
3/26/2026
MoE-Sieve: Routing-Guided LoRA for Efficient MoE Fine-Tuning
MoE-Sieve: routing-guided LoRA fine-tuning framework for MoE models that adapts to skewed expert routing patterns for efficiency.
3/26/2026
The impact of sensor placement on graph-neural-network-based leakage detection
Investigates optimal sensor placement for GNN-based leakage detection in water distribution networks.
3/26/2026
Towards Effective Experiential Learning: Dual Guidance for Utilization and Internalization
Dual guidance approach for RL-based LLM training combining external verification and internal experience to improve reasoning task performance.
3/26/2026
Causality-Driven Disentangled Representation Learning in Multiplex Graphs
Causal inference framework for learning disentangled representations from multiplex graphs by separating shared and layer-specific information.
3/26/2026
The Alignment Tax: Response Homogenization in Aligned LLMs and Its Implications for Uncertainty Estimation
RLHF-aligned LLMs exhibit response homogenization limiting uncertainty estimation; analyzes alignment tax impact across different tasks and sampling methods.
3/26/2026
On Gossip Algorithms for Machine Learning with Pairwise Objectives
Gossip-based distributed machine learning algorithms for IoT networks with privacy constraints and limited computation/communication resources.
3/26/2026
Reservoir-Based Graph Convolutional Networks
Graph convolutional networks using reservoir computing to address challenges with complex and dynamic graph data and long-range dependencies.
3/26/2026
Efficient Controller Learning from Human Preferences and Numerical Data Via Multi-Modal Surrogate Models
Bayesian optimization framework for tuning control policies using human preferences and pairwise comparisons instead of quantitative evaluations.
3/26/2026
TsetlinWiSARD: On-Chip Training of Weightless Neural Networks using Tsetlin Automata on FPGAs
FPGA-based implementation of weightless neural networks using Tsetlin automata for on-chip training and inference with low latency and complexity.
3/26/2026
A Deep Dive into Scaling RL for Code Generation with Synthetic Data and Curricula
Scalable RL pipeline for improving LLM code generation through synthetic data and curriculum learning, addressing data diversity challenges at scale.
3/26/2026
IPatch: A Multi-Resolution Transformer Architecture for Robust Time-Series Forecasting
Transformer architecture for multivariate time series forecasting using multi-resolution representations to capture short-term and long-range dependencies.
3/26/2026
Large Language Model Guided Incentive Aware Reward Design for Cooperative Multi-Agent Reinforcement Learning
Framework uses LLMs to automatically design reward functions for cooperative multi-agent reinforcement learning, synthesizing executable reward programs from environment instrumentation.
3/26/2026
CoordLight: Learning Decentralized Coordination for Network-Wide Traffic Signal Control
Multi-agent reinforcement learning approach for decentralized adaptive traffic signal control using learned coordination in partially observable environments.
3/26/2026
MolEvolve: LLM-Guided Evolutionary Search for Interpretable Molecular Optimization
MolEvolve framework uses LLM guidance with evolutionary search for interpretable molecular optimization, addressing activity cliffs and lack of interpretability.
3/26/2026
CUA-Suite: Massive Human-annotated Video Demonstrations for Computer-Use Agents
CUA-Suite dataset provides massive human-annotated continuous video demonstrations for training computer-use agents on desktop automation tasks, addressing data bottleneck.
3/26/2026
Towards Safe Learning-Based Non-Linear Model Predictive Control through Recurrent Neural Network Modeling
Sequential-AMPC uses recurrent neural networks to approximate nonlinear model predictive control offline, reducing online computation for embedded hardware control systems.
3/26/2026
Claudini: Autoresearch Discovers State-of-the-Art Adversarial Attack Algorithms for LLMs
AI agents using Claude Code autonomously discovered novel adversarial attack algorithms for LLMs that outperform 30+ existing methods in jailbreaking and prompt injection attacks.
3/26/2026
AVO: Agentic Variation Operators for Autonomous Evolutionary Search
Agentic Variation Operators replace fixed mutation/crossover in evolutionary search with autonomous coding agents consulting lineage and domain knowledge.
3/26/2026
TuneShift-KD: Knowledge Distillation and Transfer for Fine-tuned Models
TuneShift-KD enables knowledge distillation and transfer of fine-tuned specialized knowledge to newer LLM architectures without access to original training data.
3/26/2026
No Single Metric Tells the Whole Story: A Multi-Dimensional Evaluation Framework for Uncertainty Attributions
Multi-dimensional evaluation framework for uncertainty attribution methods in explainable AI addressing inconsistent evaluation across heterogeneous proxy tasks.
3/26/2026
UI-Voyager: A Self-Evolving GUI Agent Learning via Failed Experience
UI-Voyager is a self-evolving mobile GUI agent using rejection fine-tuning and credit assignment to learn from failed trajectories in long-horizon tasks.
3/26/2026
Scaling Recurrence-aware Foundation Models for Clinical Records via Next-Visit Prediction
RAVEN applies generative pretraining to structured electronic health records using recurrence-aware next-visit event prediction on 1M+ patient dataset.
3/26/2026
DreamerAD: Efficient Reinforcement Learning via Latent World Model for Autonomous Driving
DreamerAD enables efficient RL for autonomous driving via latent world model achieving 80x speedup by compressing diffusion sampling from 100 to 1 step.
3/26/2026
Polynomial Speedup in Diffusion Models with the Multilevel Euler-Maruyama Method
Multilevel Euler-Maruyama method accelerates diffusion model solving via multi-level approximators with polynomial speedup in HTMC regime.
3/26/2026
KARMA: Knowledge-Action Regularized Multimodal Alignment for Personalized Search at Taobao
KARMA applies LLMs to personalized search at Taobao by addressing knowledge-action gap through regularized multimodal alignment for next-item prediction.
3/26/2026
Beyond Masks: Efficient, Flexible Diffusion Language Models via Deletion-Insertion Processes
Deletion-Insertion Diffusion language models replace masking paradigm with discrete diffusion processes for improved computational efficiency and generation flexibility.
3/26/2026
DepthCharge: A Domain-Agnostic Framework for Measuring Depth-Dependent Knowledge in Large Language Models
DepthCharge framework measures knowledge depth in LLMs through adaptive probing across domains, addressing inability to sustain accurate responses in domain-specific details.
3/26/2026
Did You Forget What I Asked? Prospective Memory Failures in Large Language Models
Study of prospective memory failures in LLMs when formatting constraints conflict with complex tasks.
3/26/2026
MDKeyChunker: Single-Call LLM Enrichment with Rolling Keys and Key-Based Restructuring for High-Accuracy RAG
MDKeyChunker: Structure-aware document chunking and single-call LLM enrichment for improved RAG pipelines.
3/26/2026
Not All Pretraining are Created Equal: Threshold Tuning and Class Weighting for Imbalanced Polarization Tasks in Low-Resource Settings
Transformer-based approach for polarization detection in social media using threshold tuning and class weighting.
3/26/2026
LLMORPH: Automated Metamorphic Testing of Large Language Models
LLMORPH: Automated metamorphic testing tool for LLMs using metamorphic relations to verify correctness without labeled test data.
3/26/2026
LLM Inference at the Edge: Mobile, NPU, and GPU Performance Efficiency Trade-offs Under Sustained Load
Benchmark of Qwen 2.5 1.5B quantized LLM inference across mobile, NPU, and GPU platforms measuring throughput and efficiency trade-offs.
3/26/2026
Energy Efficient Software Hardware CoDesign for Machine Learning: From TinyML to Large Language Models
Comprehensive review of energy-efficient software-hardware codesign for ML from TinyML to LLMs, addressing memory and data movement bottlenecks.
3/26/2026
Dual-Gated Epistemic Time-Dilation: Autonomous Compute Modulation in Asynchronous MARL
Dual-gated approach for autonomous compute modulation in asynchronous multi-agent reinforcement learning on edge devices.
3/26/2026
Sparse Autoencoders for Interpretable Medical Image Representation Learning
Using sparse autoencoders to replace opaque vision foundation model representations with human-interpretable features for medical imaging.
3/26/2026
Object Search in Partially-Known Environments via LLM-informed Model-based Planning and Prompt Selection
LLM-informed planning framework for object search in partially-known environments using LLM probability estimates and prompt selection.
3/26/2026
Perturbation: A simple and efficient adversarial tracer for representation learning in language models
Perturbation-based method to trace and analyze linguistic representations in deep language models without imposing linearity constraints.
3/26/2026
How Vulnerable Are Edge LLMs?
Security analysis of quantized edge-deployed LLMs showing knowledge extraction attacks remain effective despite quantization noise.
3/26/2026
The DeepXube Software Package for Solving Pathfinding Problems with Learned Heuristic Functions and Search
DeepXube: Open-source Python package combining deep reinforcement learning and heuristic search to automate pathfinding problem solving.
3/26/2026
Praxium: Diagnosing Cloud Anomalies with AI-based Telemetry and Dependency Analysis
Praxium: AI-based system for diagnosing microservice anomalies in cloud applications using telemetry and dependency analysis.
3/26/2026
Self-Distillation for Multi-Token Prediction
MTP-D: Self-distillation method to improve multi-token prediction in LLMs, addressing acceptance rates and joint training challenges for faster inference.
3/26/2026
Attention-aware Inference Optimizations for Large Vision-Language Models with Memory-efficient Decoding
AttentionPack optimizes vision-language model inference with memory-efficient decoding for long sequences.
3/26/2026
ORACLE: Orchestrate NPC Daily Activities using Contrastive Learning with Transformer-CVAE
ORACLE orchestrates NPC daily activities in digital environments using contrastive learning with Transformer-CVAE.
3/26/2026
Dialogue to Question Generation for Evidence-based Medical Guideline Agent Development
LLM-based ambient assistant for evidence-based medical guidelines that surfaces targeted questions during physician consultations.
3/26/2026
The Price Reversal Phenomenon: When Cheaper Reasoning Models End Up Costing More
Systematic study reveals pricing reversal phenomenon where cheaper reasoning LLMs often cost more in practice across diverse tasks.
3/26/2026
Machine vision with small numbers of detected photons per inference
End-to-end optimized machine vision system for low-light scenarios with minimal detected photons per inference.
3/26/2026
COVTrack++: Learning Open-Vocabulary Multi-Object Tracking from Continuous Videos via a Synergistic Paradigm
COVTrack++ enables multi-object tracking for open-vocabulary categories including unseen objects using continuous video data.
3/26/2026
Minimal Sufficient Representations for Self-interpretable Deep Neural Networks
DeepIn framework for self-interpretable neural networks that identifies minimal representations needed for DNN expressiveness.
3/26/2026
Knowledge-Guided Manipulation Using Multi-Task Reinforcement Learning
KG-M3PO framework combines knowledge graphs, vision, and reinforcement learning for multi-task robotic manipulation with online 3D scene graphs.