3/4/2026
Safety Training Persists Through Helpfulness Optimization in LLM Agents
Study of safety training persistence in multi-step agentic LLM settings when optimizing for helpfulness, comparing DPO effects.
3/4/2026
Generalized Discrete Diffusion with Self-Correction
Generalized discrete diffusion model with self-correction during pretraining using uniform-absorbing objective.
3/4/2026
Beyond Binary Preferences: A Principled Framework for Reward Modeling with Ordinal Feedback
Principled mathematical framework for reward modeling leveraging ordinal preference feedback from human annotators for LLM alignment.
3/4/2026
Adaptive Personalized Federated Learning via Multi-task Averaging of Kernel Mean Embeddings
Personalized federated learning approach using kernel mean embeddings to learn inter-agent weight combinations without raw data sharing.
3/4/2026
Talking with Verifiers: Automatic Specification Generation for Neural Network Verification
Framework for automatic specification generation to improve neural network verification tool adoption by supporting higher-level semantic constraints.
3/4/2026
CUDABench: Benchmarking LLMs for Text-to-CUDA Generation
CUDABench benchmark for evaluating LLM text-to-CUDA code generation with performance assessment metrics for GPU kernels.
3/4/2026
Concept Heterogeneity-aware Representation Steering
Method for steering LLM behavior via representation manipulation that accounts for heterogeneous concept encoding across embedding spaces.
3/4/2026
Length Generalization Bounds for Transformers
Theoretical analysis of length generalization bounds for transformers on CRASP language class, addressing model generalization guarantees.
3/4/2026
High-order Knowledge Based Network Controllability Robustness Prediction: A Hypergraph Neural Network Approach
Hypergraph neural network approach for predicting network controllability robustness against attacks, replacing computationally expensive simulations.
3/4/2026
Boosting Meta-Learning for Few-Shot Text Classification via Label-guided Distance Scaling
Label-guided distance scaling method for few-shot text classification, improving meta-learner effectiveness with selective label guidance.
3/4/2026
PRISM: Exploring Heterogeneous Pretrained EEG Foundation Model Transfer to Clinical Differential Diagnosis
PRISM foundation model for EEG diagnosis using masked autoencoders, ablated across pretraining populations and clinical adaptation domains.
3/4/2026
Graph Attention Based Prioritization of Disease Responsible Genes from Multimodal Alzheimer's Network
Graph transformer framework (NETRA) for prioritizing disease genes in Alzheimer's networks using multimodal biological data.
3/4/2026
A Comparative Study of UMAP and Other Dimensionality Reduction Methods
Comparative analysis of UMAP with PCA, Kernel PCA, SIR variants, and t-SNE for dimensionality reduction across benchmarks.
3/4/2026
Temporal Imbalance of Positive and Negative Supervision in Class-Incremental Learning
Addressing catastrophic forgetting in class-incremental learning by analyzing temporal imbalance of positive/negative supervision signals.
3/4/2026
Quantum-Inspired Fine-Tuning for Few-Shot AIGC Detection via Phase-Structured Reparameterization
Quantum-enhanced LoRA fine-tuning method for few-shot AI-generated content detection, combining quantum neural networks with low-rank adaptation.
3/4/2026
Preconditioned Score and Flow Matching
Preconditioning techniques for flow matching and score-based diffusion to improve optimization by handling ill-conditioned covariance matrices.
3/4/2026
Diffusion-MPC in Discrete Domains: Feasibility Constraints, Horizon Effects, and Critic Alignment: Case study with Tetris
Diffusion-based model predictive control with discrete denoising for game playing, tested on Tetris with feasibility constraints and critic alignment.
3/4/2026
Learning graph topology from metapopulation epidemic encoder-decoder
Joint inference of epidemic parameters and mobility networks in metapopulation models using encoder-decoder architecture.
3/4/2026
Learning Optimal Search Strategies
Learning optimal threshold-based stopping rules for parking problems with unknown Poisson arrival processes via jump intensity estimation.
3/4/2026
Rigidity-Aware Geometric Pretraining for Protein Design and Conformational Ensembles
Proposes rigidity-aware geometric pretraining for protein design and conformational ensembles using global geometric representations.
3/4/2026
Personalized Multi-Agent Average Reward TD-Learning via Joint Linear Approximation
Studies personalized multi-agent average reward TD learning with joint linear approximation, inspired by federated learning approaches.
3/4/2026
A Unified Revisit of Temperature in Classification-Based Knowledge Distillation
Analyzes temperature parameter selection in knowledge distillation and its interaction with optimizer, pretraining, and finetuning choices.
3/4/2026
Spectral Regularization for Diffusion Models
Introduces loss-level spectral regularization using Fourier and wavelet-domain losses to improve diffusion model training without architecture changes.
3/4/2026
Can Computational Reducibility Lead to Transferable Models for Graph Combinatorial Optimization?
Studies computational reducibility in neural solvers for graph combinatorial optimization, enabling model generalization across task distributions.
3/4/2026
MUSE: A Run-Centric Platform for Multimodal Unified Safety Evaluation of Large Language Models
MUSE is an open-source platform for multimodal safety evaluation of LLMs with cross-modal payload generation and multi-turn attack algorithms.
3/4/2026
What Capable Agents Must Know: Selection Theorems for Robust Decision-Making under Uncertainty
Proves selection theorems showing that low average-case regret forces AI agents to develop internal world models or belief states for robust decision-making.
3/4/2026
ParEVO: Synthesizing Code for Irregular Data: High-Performance Parallelism through Agentic Evolution
ParEVO uses LLM-based agentic evolution to synthesize parallel code for irregular data structures, addressing limitations of standard models on concurrent programming.
3/4/2026
Thermodynamic Regulation of Finite-Time Gibbs Training in Energy-Based Models: A Restricted Boltzmann Machine Study
Studies thermodynamic regulation of finite-time Gibbs chain training in Restricted Boltzmann Machines, analyzing energy landscape evolution during learning.
3/4/2026
Bridging Diffusion Guidance and Anderson Acceleration via Hopfield Dynamics
Establishes theoretical connection between classifier-free guidance in diffusion models and Anderson acceleration via Hopfield dynamics.
3/4/2026
EdgeFLow: Serverless Federated Learning via Sequential Model Migration in Edge Networks
Presents EdgeFLow, a federated learning framework using sequential model migration in edge networks to reduce communication bottlenecks in IoT systems.
3/4/2026
Wasserstein Proximal Policy Gradient
Derives Wasserstein Proximal Policy Gradient using optimal transport geometry for continuous-action entropy-regularized RL without policy log-density evaluation.
3/4/2026
Towards Parameter-Free Temporal Difference Learning
Develops parameter-free temporal difference learning for RL that avoids requiring problem-dependent quantities like feature covariance eigenvalues.
3/4/2026
Joint Optimization of Model Partitioning and Resource Allocation for Anti-Jamming Collaborative Inference Systems
Studies DNN partitioning and resource allocation for device-edge collaborative inference under jamming attacks on resource-constrained systems.
3/4/2026
Heterogeneous Agent Collaborative Reinforcement Learning
Introduces HACRL, a collaborative reinforcement learning paradigm where heterogeneous agents share verified rollouts during training but execute independently at inference.
3/4/2026
Real-Time Generative Policy via Langevin-Guided Flow Matching for Autonomous Driving
Proposes diffusion actor-critic with flow matching for real-time autonomous driving policies, addressing inference latency in generative RL approaches.
3/4/2026
Same Error, Different Function: The Optimizer as an Implicit Prior in Financial Time Series
Neural networks on financial time series show underspecification where different optimizers produce identical test loss but learn different functions in volatility forecasting.
3/4/2026
Implicit Bias in Deep Linear Discriminant Analysis
Theoretical analysis of implicit regularization in Deep Linear Discriminant Analysis for metric learning objectives.
3/4/2026
Post Hoc Extraction of Pareto Fronts for Continuous Control
Multi-objective reinforcement learning method for extracting Pareto fronts of policies in continuous control tasks, addressing trade-offs between multiple objectives.
3/4/2026
MASPOB: Bandit-Based Prompt Optimization for Multi-Agent Systems with Graph Neural Networks
Bandit-based prompt optimization for multi-agent systems using graph neural networks to improve LLM-powered workflow performance without modifying workflows.
3/4/2026
Robust Heterogeneous Analog-Digital Computing for Mixture-of-Experts Models with Theoretical Generalization Guarantees
Heterogeneous analog-digital computing approach for efficient Mixture-of-Experts inference with theoretical generalization guarantees and hardware nonideality mitigation.
3/4/2026
SaFeR-ToolKit: Structured Reasoning via Virtual Tool Calling for Multimodal Safety
SaFeR-ToolKit formalizes multimodal safety as checkable protocol using virtual tool calling for vision-language models to prevent jailbreaks.
3/4/2026
HomeAdam: Adam and AdamW Algorithms Sometimes Go Home to Obtain Better Provable Generalization
HomeAdam variant of Adam optimizer improves generalization bounds to match SGD convergence rates for deep learning model training.
3/4/2026
Improving Diffusion Planners by Self-Supervised Action Gating with Energies
SAGE method improves diffusion planners for offline RL by using latent consistency signals to penalize dynamically inconsistent plans at inference-time.
3/4/2026
From Shallow to Deep: Pinning Semantic Intent via Causal GRPO
Two-Stage Causal-GRPO framework addresses shallow safety alignment in LLMs vulnerable to adversarial prefix attacks through semantic intent pinning.
3/4/2026
Causal Learning Should Embrace the Wisdom of the Crowd
Paradigm for causal structure learning from observational data leveraging human causal knowledge to address combinatorial explosion of possible graphs.
3/4/2026
Addressing Missing and Noisy Modalities in One Solution: Unified Modality-Quality Framework for Low-quality Multimodal Data
Unified framework addressing both missing and noisy modalities in multimodal learning to improve robustness on low-quality real-world data.
3/4/2026
An Empirical Analysis of Calibration and Selective Prediction in Multimodal Clinical Condition Classification
Empirical evaluation of uncertainty-based selective prediction reliability in multimodal clinical condition classification using ICU data.
3/4/2026
The power of small initialization in noisy low-tubal-rank tensor recovery
Theoretical analysis of factorized gradient descent for low-tubal-rank tensor recovery from noisy linear measurements under t-product framework.
3/4/2026
Practical FP4 Training for Large-Scale MoE Models on Hopper GPUs
Training recipe enabling FP4 efficiency for large-scale Mixture-of-Experts models on Hopper GPUs without native 4-bit support.
3/4/2026
Deep learning-guided evolutionary optimization for protein design
BoGA framework combines evolutionary search with Bayesian optimization for protein sequence design and function prediction.