2/24/2026
From Few-Shot to Zero-Shot: Towards Generalist Graph Anomaly Detection
Proposes generalist graph anomaly detection method enabling zero-shot transfer across datasets without dataset-specific training.
2/24/2026
Bayesian Lottery Ticket Hypothesis
Applies lottery ticket hypothesis to Bayesian neural networks, finding sparse subnetworks for uncertainty quantification with reduced computational cost.
2/24/2026
L2G-Net: Local to Global Spectral Graph Neural Networks via Cauchy Factorizations
Spectral graph neural networks via Cauchy factorizations enabling local-to-global spectral methods without expensive eigenbasis computation.
2/24/2026
Rank-Aware Spectral Bounds on Attention Logits for Stable Low-Precision Training
Rank-aware concentration inequality for attention logit magnitudes in transformers, improving stability of low-precision training via spectral bounds.
2/24/2026
Issues with Measuring Task Complexity via Random Policies in Robotic Tasks
Analysis of task complexity measurement via random policies in non-tabular RL domains, examining statistical and information-theoretic metrics.
2/24/2026
VariBASed: Variational Bayes-Adaptive Sequential Monte-Carlo Planning for Deep Reinforcement Learning
Variational Bayes-adaptive planning for deep reinforcement learning, balancing exploration-exploitation with scalable belief-state computation.
2/24/2026
Boosting for Vector-Valued Prediction and Conditional Density Estimation
Theoretical analysis of boosting for vector-valued prediction and conditional density estimation under general divergences and stability conditions.
2/24/2026
PCA-VAE: Differentiable Subspace Quantization without Codebook Collapse
PCA-VAE replaces vector quantization with differentiable online PCA bottleneck via Oja's rule, eliminating codebook collapse and straight-through estimators.
2/24/2026
TRUE: A Trustworthy Unified Explanation Framework for Large Language Model Reasoning
Trustworthy Unified Explanation framework for interpreting LLM reasoning, revealing stability and systematic failure mechanisms across instances.
2/24/2026
DeepInterestGR: Mining Deep Multi-Interest Using Multi-Modal LLMs for Generative Recommendation
Generative recommendation framework using multi-modal LLMs to mine deep multi-interests beyond shallow behavioral signals for semantic ID prediction.
2/24/2026
From Human-Level AI Tales to AI Leveling Human Scales
Framework for calibrating AI benchmark performance against world population baselines to provide human-anchored capability scales.
2/24/2026
LoMime: Query-Efficient Membership Inference using Model Extraction in Label-Only Settings
Membership inference attacks on ML models using model extraction in label-only settings without access to confidence scores or shadow models.
2/24/2026
Exponential Convergence of (Stochastic) Gradient Descent for Separable Logistic Regression
Theoretical analysis of gradient descent convergence rates for separable logistic regression under large step sizes and unstable regimes.
2/24/2026
Why ReLU? A Bit-Model Dichotomy for Deep Network Training
Theoretical analysis of neural network training complexity under Real-RAM vs bit-model computation, proving ERM for simple networks is ∃ℝ-complete.
2/24/2026
Learning to Detect Language Model Training Data via Active Reconstruction
Proposes Active Data Reconstruction Attack (ADRA) for detecting LLM training data through active model manipulation rather than passive membership inference.
2/24/2026
Pushing the Limits of Inverse Lithography with Generative Reinforcement Learning
Applies generative RL to inverse lithography for semiconductor manufacturing mask synthesis, replacing deterministic approaches with conditional sampling.
2/24/2026
A Markovian View of Iterative-Feedback Loops in Image Generative Models: Neural Resonance and Model Collapse
Analyzes model collapse in image generative models through iterative feedback loops using Markovian framework, revealing neural resonance phenomena in latent space.
2/24/2026
Back to Blackwell: Closing the Loop on Intransitivity in Multi-Objective Preference Fine-Tuning
Addresses intransitive preferences in LLM fine-tuning via preference learning, proposing methods to handle cyclic preference conflicts in multi-objective optimization.
2/24/2026
IDLM: Inverse-distilled Diffusion Language Models
Inverse distillation for diffusion language models. Accelerates discrete diffusion models for faster text generation inference.
2/24/2026
Learning from Complexity: Exploring Dynamic Sample Pruning of Spatio-Temporal Training
Dynamic sample pruning for spatio-temporal forecasting. Optimizes training data efficiency for deep learning on large datasets.
2/24/2026
Robust Predictive Uncertainty and Double Descent in Contaminated Bayesian Random Features
Robust Bayesian random feature regression with contaminated priors. Studies double descent phenomenon under model misspecification.
2/24/2026
Detecting labeling bias using influence functions
Influence functions for detecting labeling bias in datasets. Addresses fairness issues from biased data collection.
2/24/2026
Test-Time Learning of Causal Structure from Interventional Data
Test-time learning method for causal structure discovery from interventional data. Combines test-time training with causal inference.
2/24/2026
Celo2: Towards Learned Optimization Free Lunch
Celo2 learned optimizer with improved meta-generalization. Aims for practical adoption of learned optimization rules beyond hand-designed optimizers.
2/24/2026
Incremental Learning of Sparse Attention Patterns in Transformers
Analysis of how transformers learn sparse attention patterns incrementally. Studies information integration from multiple past positions.
2/24/2026
Virtual Parameter Sharpening: Dynamic Low-Rank Perturbations for Inference-Time Reasoning Enhancement
Virtual Parameter Sharpening for inference-time reasoning enhancement. Dynamic low-rank perturbations for transformer adaptation without persistent parameters.
2/24/2026
Online Realizable Regression and Applications for ReLU Networks
Theoretical analysis of realizable online regression under metric-like losses. Studies ReLU networks in adversarial setting.
2/24/2026
Adaptive Problem Generation via Symbolic Representations
Adaptive problem generation via symbolic representations for training small open-weight LMs on math tasks. Data generation using RL with verifiable rewards.
2/24/2026
How to Allocate, How to Learn? Dynamic Rollout Allocation and Advantage Modulation for Policy Optimization
Dynamic rollout allocation and policy optimization for LLM reasoning with verifiable rewards. Improves RL training efficiency for reasoning tasks.
2/24/2026
Understanding Empirical Unlearning with Combinatorial Interpretability
Combinatorial interpretability framework for understanding knowledge persistence in unlearning. Studies how information is retained in foundation models.
2/24/2026
Evaluating SAP RPT-1 for Enterprise Business Process Prediction: In-Context Learning vs. Traditional Machine Learning on Structured SAP Data
Evaluation of SAP's RPT-1 tabular foundation model on enterprise data. Compares in-context learning vs traditional ML on structured datasets.
2/24/2026
Alternating Bi-Objective Optimization for Explainable Neuro-Fuzzy Systems
Gradient-based optimization for explainable neuro-fuzzy systems. Addresses accuracy-explainability tradeoff in fuzzy AI models.
2/24/2026
DGPO: RL-Steered Graph Diffusion for Neural Architecture Generation
RL-steered graph diffusion for neural architecture search. Uses reinforcement learning to guide generative models for DAG-based NAS.
2/24/2026
Taming Preconditioner Drift: Unlocking the Potential of Second-Order Optimizers for Federated Learning on Non-IID Data
Addresses preconditioner drift in federated second-order optimizers on non-IID data through curvature alignment techniques.
2/24/2026
Soft Sequence Policy Optimization: Bridging GMPO and SAPO
Bridges GMPO and SAPO by combining sequence-level importance sampling with soft clipping alternatives for improved LLM policy optimization.
2/24/2026
CTS-Bench: Benchmarking Graph Coarsening Trade-offs for GNNs in Clock Tree Synthesis
Benchmarks graph coarsening trade-offs for GNNs applied to clock tree synthesis in electronic design automation.
2/24/2026
Training-Free Cross-Architecture Merging for Graph Neural Networks
Introduces H-GRAMA for merging heterogeneous GNN architectures through routing and message aggregation without retraining.
2/24/2026
Smooth Gate Functions for Soft Advantage Policy Optimization
Proposes Soft Adaptive Policy Optimization (SAPO) replacing hard clipping with smooth sigmoid gate functions to stabilize LLM training and reasoning in GRPO framework.
2/24/2026
Active perception and disentangled representations allow continual, episodic zero and few-shot learning
Framework combining active perception and disentangled representations for continual and few-shot learning without destructive interference.
2/24/2026
LLMs Can Learn to Reason Via Off-Policy RL
Method for training LLMs to reason using off-policy reinforcement learning, addressing policy lag in distributed training architectures.
2/24/2026
Stable Deep Reinforcement Learning via Isotropic Gaussian Representations
Analysis showing isotropic Gaussian representations improve stability in deep RL under non-stationary training dynamics.
2/24/2026
Spiking Graph Predictive Coding for Reliable OOD Generalization
Spiking graph neural network with predictive coding framework for out-of-distribution generalization in dynamic web environments.
2/24/2026
Federated Causal Representation Learning in State-Space Systems for Decentralized Counterfactual Reasoning
Federated learning approach for causal representation learning in state-space systems enabling decentralized counterfactual reasoning across networked assets.
2/24/2026
SenTSR-Bench: Thinking with Injected Knowledge for Time-Series Reasoning
Benchmark combining general reasoning LLMs with domain-specific time-series knowledge for improved time-series diagnostic reasoning tasks.
2/24/2026
Making Conformal Predictors Robust in Healthcare Settings: a Case Study on EEG Classification
Evaluation of conformal prediction methods for EEG classification handling distribution shifts in healthcare without standard i.i.d. assumptions.
2/24/2026
Federated Learning Playground
Interactive browser-based educational platform for learning Federated Learning concepts with real-time visualization of heterogeneous data effects.
2/24/2026
Grokking Finite-Dimensional Algebra
Investigation of grokking phenomenon in neural networks learning multiplication in finite-dimensional algebras beyond group operations.
2/24/2026
The Sample Complexity of Replicable Realizable PAC Learning
Theoretical analysis of sample complexity bounds in replicable realizable PAC learning using Cayley graphs and spectral analysis.
2/24/2026
Leap+Verify: Regime-Adaptive Speculative Weight Prediction for Accelerating Neural Network Training
Leap+Verify applies speculative execution to accelerate neural network training by predicting and validating future weights across detected regimes.
2/24/2026
Interpolation-Driven Machine Learning Approaches for Plume Shine Dose Estimation: A Comparison of XGBoost, Random Forest, and TabNet
Compares XGBoost, Random Forest, and TabNet for radiation dose estimation in nuclear safety using interpolation-driven ML approaches.