3/13/2026
On the Reliability of Cue Conflict and Beyond
Analysis of cue-conflict benchmarks for measuring neural network shape-texture bias, identifying instability in stylization-based bias estimation methods.
3/13/2026
Historical Consensus: Preventing Posterior Collapse via Iterative Selection of Gaussian Mixture Priors
Historical Consensus approach for preventing posterior collapse in VAEs through iterative Gaussian mixture prior selection based on data covariance spectral properties.
3/13/2026
Structure-Aware Epistemic Uncertainty Quantification for Neural Operator PDE Surrogates
Uncertainty quantification framework for neural operator PDE surrogates emphasizing computational efficiency and spatial localization of epistemic uncertainty.
3/13/2026
Interventional Time Series Priors for Causal Foundation Models
Interventional time series data generator for training causal foundation models, extending prior-data fitted networks to time series with ground-truth interventions.
3/13/2026
Graph Tokenization for Bridging Graphs and Transformers
Graph tokenization framework combining reversible serialization with BPE to enable transformer models to process graph-structured data as token sequences.
3/13/2026
Task-Conditioned Routing Signatures in Sparse Mixture-of-Experts Transformers
Analysis of routing signatures in sparse Mixture-of-Experts transformers to understand task-conditioned expert selection patterns in large language models.
3/13/2026
Learning Tree-Based Models with Gradient Descent
Gradient descent-based approach for learning interpretable tree-based decision models, addressing combinatorial complexity of traditional discrete tree learning.
3/13/2026
A Learning-Based Superposition Operator for Non-Renewal Arrival Processes in Queueing Networks
Data-driven superposition operator for non-renewal arrival processes in queueing networks using moment-based learning instead of classical analytical methods.
3/13/2026
High-resolution weather-guided surrogate modeling for data-efficient cross-location building energy prediction
Surrogate modeling approach for building energy prediction using weather-guided models to reduce computational cost of physics-based EnergyPlus simulations.
3/13/2026
Higher-Order Modular Attention: Fusing Pairwise and Triadic Interactions for Protein Sequences
Higher-Order Modular Attention (HOMA) mechanism extending transformer attention to capture triadic interactions in protein sequences beyond pairwise dependencies.
3/13/2026
Scaling Reasoning Efficiently via Relaxed On-Policy Distillation
REOPOLD: framework for on-policy distillation of reasoning capabilities to smaller models using policy optimization and token-level rewards for efficient scaling.
3/13/2026
H2LooP Spark Preview: Continual Pretraining of Large Language Models for Low-Level Embedded Systems Code
H2LooP Spark Preview: continual pretraining framework for LLMs specialized in low-level embedded systems code generation with hardware-specific domain adaptation.
3/13/2026
Attention Gathers, MLPs Compose: A Causal Analysis of an Action-Outcome Circuit in VideoViT
Mechanistic interpretability analysis of video vision transformers to understand how internal circuits represent action-outcome relationships in classification tasks.
3/13/2026
Systematic Scaling Analysis of Jailbreak Attacks in Large Language Models
Scaling-law framework for jailbreak attacks on LLMs treating each attack as compute-bounded optimization across methods and model families.
3/13/2026
Algorithmic Capture, Computational Complexity, and Inductive Bias of Infinite Transformers
Formal analysis of algorithmic learning (grokking) in infinite-width transformers with computational complexity bounds and generalization theory.
3/13/2026
Bayesian Optimization of Partially Known Systems using Hybrid Models
Bayesian optimization method combining probabilistic models with hybrid systems for efficient black-box optimization of expensive functions.
3/13/2026
Representation Finetuning for Continual Learning
Parameter-efficient fine-tuning approach for continual learning that controls representation-level adaptation in pretrained models.
3/13/2026
Reference-Guided Machine Unlearning
Machine unlearning method using reference-guided approach to remove data influence from trained models while preserving utility.
3/13/2026
Client-Conditional Federated Learning via Local Training Data Statistics
Federated learning method conditioning global model on client-specific PCA statistics for heterogeneous data without extra communication.
3/13/2026
Hindsight-Anchored Policy Optimization: Turning Failure into Feedback in Sparse Reward Settings
Hindsight-Anchored Policy Optimization for sparse-reward RL in reasoning model post-training, addressing advantage collapse.
3/13/2026
Meta-Reinforcement Learning with Self-Reflection for Agentic Search
MR-Search: meta-reinforcement learning framework for agentic search with self-reflection, enabling agents to improve in-context exploration.
3/13/2026
Jailbreak Scaling Laws for Large Language Models: Polynomial-Exponential Crossover
Analysis of adversarial attacks on LLMs showing polynomial-exponential crossover in success rates with inference-time samples.
3/13/2026
abx_amr_simulator: A simulation environment for antibiotic prescribing policy optimization under antimicrobial resistance
Open-source Python simulation package for modeling antibiotic prescribing and antimicrobial resistance dynamics in RL-compatible environment.
3/13/2026
ZTab: Domain-based Zero-shot Annotation for Table Columns
Zero-shot learning approach for automatically detecting semantic column types in relational tables without labeled training data.
3/13/2026
UniHetCO: A Unified Heterogeneous Representation for Multi-Problem Learning in Unsupervised Neural Combinatorial Optimization
Unsupervised neural combinatorial optimization method (UniHetCO) for multi-problem learning on graph subset-selection problems without ground-truth solutions.
3/13/2026
Bridging Discrete Marks and Continuous Dynamics: Dual-Path Cross-Interaction for Marked Temporal Point Processes
Dual-path approach combining discrete mark prediction and continuous dynamics for marked temporal point processes using neural ODEs.
3/13/2026
Slack More, Predict Better: Proximal Relaxation for Probabilistic Latent Variable Model-based Soft Sensors
Proximal relaxation method for improving nonlinear probabilistic latent variable models in soft sensor applications, addressing training inefficiencies.
3/13/2026
Deep Learning Network-Temporal Models For Traffic Prediction
Deep learning network-temporal models for multivariate time series traffic prediction, addressing topological interdependency and complex temporal patterns.
3/13/2026
Leveraging Phytolith Research using Artificial Intelligence
Sorometry pipeline for automated phytolith analysis using AI to digitize and classify microscope images, replacing manual labor-intensive analysis.
3/13/2026
Grammar of the Wave: Towards Explainable Multivariate Time Series Event Detection via Neuro-Symbolic VLM Agents
Neuro-symbolic VLM agents for time series event detection using natural language descriptions, addressing semantic event classification with scarce labeled data.
3/13/2026
Attention Sinks Are Provably Necessary in Softmax Transformers: Evidence from Trigger-Conditional Tasks
Theoretical analysis proving attention sinks are necessary in softmax transformers for trigger-conditional tasks, formalizing why attention collapses to content-agnostic positions.
3/13/2026
KEPo: Knowledge Evolution Poison on Graph-based Retrieval-Augmented Generation
KEPo analyzes poisoning attacks on Graph-RAG systems where attackers inject malicious texts to manipulate LLM outputs.
3/13/2026
Sharpness-Aware Minimization for Generalized Embedding Learning in Federated Recommendation
Sharpness-aware minimization for stable item embedding learning in federated recommendation systems preserving privacy.
3/13/2026
LongFlow: Efficient KV Cache Compression for Reasoning M
LongFlow compresses KV cache for reasoning models like o1 and R1, reducing memory and bandwidth during long output generation.
3/13/2026
CFD-HAR: User-controllable Privacy through Conditional Feature Disentanglement
Conditional feature disentanglement approach for user-controllable privacy in wearable sensor-based human activity recognition.
3/13/2026
Multi-Task Anti-Causal Learning for Reconstructing Urban Events from Residents' Reports
Multi-Task Anti-Causal learning framework exploiting cross-task invariances to infer latent causes from observed urban event reports.
3/13/2026
CAETC: Causal Autoencoding and Treatment Conditioning for Counterfactual Estimation over Time
CAETC method using adversarial autoencoding for counterfactual estimation with time-dependent confounding in observational data.
3/13/2026
Survival Meets Classification: A Novel Framework for Early Risk Prediction Models of Chronic Diseases
Integrates survival analysis with classification for early chronic disease risk prediction using EMR data.
3/13/2026
Hybrid Energy-Aware Reward Shaping: A Unified Lightweight Physics-Guided Methodology for Policy Optimization
H-EARS combines potential-based reward shaping with energy-aware regularization for efficient deep reinforcement learning control.
3/13/2026
AutoScout: Structured Optimization for Automating ML System Configuration
AutoScout automates ML system configuration via structured optimization over model parallelism, communication, and runtime parameters.
3/13/2026
Fractional Rotation, Full Potential? Investigating Performance and Convergence of Partial RoPE
Investigates partial RoPE rotations in transformers, reducing memory at long contexts while maintaining performance.
3/13/2026
Personalized Federated Learning via Gaussian Generative Modeling
Personalized federated learning using Gaussian generative modeling to handle data heterogeneity across distributed clients.
3/13/2026
Simple Recipe Works: Vision-Language-Action Models are Natural Continual Learners with Reinforcement Learning
Studies continual RL for Vision-Language-Action models, finding sequential fine-tuning avoids catastrophic forgetting without complex strategies.
3/13/2026
Context-dependent manifold learning: A neuromodulated constrained autoencoder approach
Neuromodulated constrained autoencoders for context-dependent dimensionality reduction in varying environments.
3/13/2026
Entropy-Preserving Reinforcement Learning
Policy gradient methods for LLM reasoning naturally reduce trajectory diversity; proposes entropy-preserving training approach.
3/13/2026
EvoFlows: Evolutionary Edit-Based Flow-Matching for Protein Engineering
EvoFlows model for protein engineering using edit-based flow-matching to predict mutations on template sequences.
3/13/2026
Mitigating the Multiplicity Burden: The Role of Calibration in Reducing Predictive Multiplicity of Classifiers
Examines calibration's role in reducing predictive multiplicity and improving stability in high-stakes ML classifier deployments.
3/13/2026
Exploiting Expertise of Non-Expert and Diverse Agents in Social Bandit Learning: A Free Energy Approach
Social bandit learning framework combining individual and collective learning in populations of reinforcement learning agents.
3/13/2026
A Further Efficient Algorithm with Best-of-Both-Worlds Guarantees for $m$-Set Semi-Bandit Problem
Theoretical study of Follow-the-Perturbed-Leader algorithm optimality in semi-bandit problems with best-of-both-worlds guarantees.
3/13/2026
Language Generation with Replay: A Learning-Theoretic View of Model Collapse
Analyzes model collapse when LLM-generated text re-enters training data as data consumption grows, proposing replay-based solutions.