2/16/2026
Pareto-Conditioned Diffusion Models for Offline Multi-Objective Optimization
Pareto-Conditioned Diffusion framework for offline multi-objective optimization via conditional sampling from diffusion models on static datasets.
2/16/2026
Dispelling the Curse of Singularities in Neural Network Optimizations
Analysis of optimization instability in deep neural networks caused by parametric singularities, proposing gradient Frobenius norm solutions.
2/16/2026
Exploring AI-Augmented Sensemaking of Patient-Generated Health Data: A Mixed-Method Study with Healthcare Professionals in Cardiac Risk Reduction
Study explores how LLMs support healthcare professionals in analyzing patient-generated health data from wearables and smartphones for cardiac risk reduction.
2/16/2026
AEGIS: Adversarial Target-Guided Retention-Data-Free Robust Concept Erasure from Diffusion Models
AEGIS method for concept erasure from diffusion models balances robustness against concept reactivation with retention of model utility under adversarial conditions.
2/16/2026
Principled Synthetic Data Enables the First Scaling Laws for LLMs in Recommendation
Framework establishes first scaling laws for LLMs in recommendation systems using principled synthetic data to replace noisy user interaction data in pre-training.
2/16/2026
Learnable Chernoff Baselines for Inference-Time Alignment
Learnable Chernoff Baselines enable efficient inference-time reward-guided alignment for generative models without architecture modifications or computational overhead.
2/16/2026
Bielik Guard: Efficient Polish Language Safety Classifiers for LLM Content Moderation
Bielik Guard provides efficient safety classifiers for Polish-language LLM content moderation, offering 0.1B and 0.5B parameter model variants based on RoBERTa.
2/16/2026
GISA: A Benchmark for General Information-Seeking Assistant
GISA benchmark evaluates information-seeking AI agents performing multi-turn web interactions, addressing limitations of existing benchmarks with naturally-constructed real-world tasks.
2/16/2026
LLaDA2.1: Speeding Up Text Diffusion via Token Editing
LLaDA2.1 improves text diffusion model decoding speed by combining Token-to-Token and Mask-to-Token editing schemes for 100B-parameter block-diffusion models.
2/16/2026
Predicting Open Source Software Sustainability with Deep Temporal Neural Hierarchical Architectures and Explainable AI
Deep learning model with explainability for predicting open source software project sustainability from temporal contribution patterns.
2/16/2026
Less is Enough: Synthesizing Diverse Data in Feature Space of LLMs
Method for synthesizing diverse post-training data for LLMs using Feature Activation Coverage metric based on model representations.
2/16/2026
Low-Dimensional Execution Manifolds in Transformer Learning Dynamics: Evidence from Modular Arithmetic Tasks
Investigation of learning dynamics in transformers showing training trajectories collapse onto low-dimensional manifolds in modular arithmetic tasks.
2/16/2026
Blind Gods and Broken Screens: Architecting a Secure, Intent-Centric Mobile Agent Operating System
Security analysis of mobile LLM agents identifying vulnerabilities in screen-interface paradigm and proposing intent-centric architecture.
2/16/2026
HiFloat4 Format for Language Model Inference
Block floating-point data format for efficient LLM inference achieving 4.5 bits per value with hierarchical scaling.
2/16/2026
Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation
Behavioral study examining user interaction with three LLM assistance modalities (Advisor, Coach, Delegate) in multi-party negotiation scenarios.
2/16/2026
DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing
Lightweight 5B parameter multimodal model for image generation and editing competitive with much larger models.
2/16/2026
OptiML: An End-to-End Framework for Program Synthesis and CUDA Kernel Optimization
Framework for synthesizing and optimizing CUDA kernels using LLMs combined with systematic exploration to achieve competitive hardware performance.
2/16/2026
Abstractive Red-Teaming of Language Model Character
Method for identifying queries that cause LLM character specification violations using red-teaming approaches to detect deployment-level failures efficiently.
2/16/2026
The Appeal and Reality of Recycling LoRAs with Adaptive Merging
Research on adaptively merging multiple LoRA modules from open models to improve performance on downstream tasks.
2/16/2026
Value Bonuses using Ensemble Errors for Exploration in Reinforcement Learning
Exploration method in reinforcement learning using ensemble error estimates to compute optimistic value bonuses for directed agent exploration.
2/16/2026
Synthetic Interaction Data for Scalable Personalization in Large Language Models
Addresses LLM personalization by generating synthetic user-specific interaction data at scale to optimize prompts for individual user preferences and constraints.
2/16/2026
AstRL: Analog and Mixed-Signal Circuit Synthesis with Deep Reinforcement Learning
Applies deep reinforcement learning to automate analog and mixed-signal circuit design optimization across diverse non-differentiable design spaces.
2/16/2026
Soft Contamination Means Benchmarks Test Shallow Generalization
Studies soft contamination in LLM training data through semantic duplicates, showing typical decontamination filters fail to detect near-equivalent benchmark test data.
2/16/2026
Stabilizing Native Low-Rank LLM Pretraining
Demonstrates stable training of LLMs from scratch using exclusively low-rank weight factorization, matching dense model performance while reducing computational costs.
2/16/2026
Safe Reinforcement Learning via Recovery-based Shielding with Gaussian Process Dynamics Models
Safe reinforcement learning framework using recovery-based shielding with Gaussian process models for non-linear continuous control with provable safety guarantees.
2/16/2026
Continuous Diffusion Models Can Obey Formal Syntax
Training-free guidance method enabling continuous diffusion language models to satisfy formal syntactic constraints like JSON schema matching via regular expressions.
2/16/2026
Regularized Meta-Learning for Improved Generalization
Regularized meta-learning framework addressing redundancy and overfitting in deep ensemble methods through redundancy-aware projection and statistical weighting.
2/16/2026
Designing RNAs with Language Models
RNA sequence design reframed as conditional sequence generation task using language models instead of traditional optimization approaches.
2/16/2026
A Theoretical Analysis of Mamba's Training Dynamics: Filtering Relevant Features for Generalization in State Space Models
Theoretical analysis of Mamba state space model training dynamics and generalization properties.
2/16/2026
On Robustness and Chain-of-Thought Consistency of RL-Finetuned VLMs
Analysis of robustness and reasoning consistency in vision language models fine-tuned with reinforcement learning.
2/16/2026
Bench-MFG: A Benchmark Suite for Learning in Stationary Mean Field Games
Bench-MFG: standardized benchmark suite for mean field games and multi-agent reinforcement learning.
2/16/2026
Multi-Agent Model-Based Reinforcement Learning with Joint State-Action Learned Embeddings
Multi-agent model-based RL with joint state-action learned embeddings for coordination in dynamic environments.
2/16/2026
Constraint-Rectified Training for Efficient Chain-of-Thought
Constraint-rectified training method to reduce overhead and overthinking in chain-of-thought reasoning.
2/16/2026
Flow-Factory: A Unified Framework for Reinforcement Learning in Flow-Matching Models
Flow-Factory: unified modular framework for reinforcement learning with flow-matching and diffusion models.
2/16/2026
AMPS: Adaptive Modality Preference Steering via Functional Entropy
Adaptive steering method to balance modality preferences in multimodal large language models.
2/16/2026
Exploring Accurate and Transparent Domain Adaptation in Predictive Healthcare via Concept-Grounded Orthogonal Inference
Concept-grounded transparent domain adaptation for clinical event prediction on electronic health records.
2/16/2026
Fractional Order Federated Learning for Battery Electric Vehicle Energy Consumption Modeling
Fractional-order federated learning for battery electric vehicle energy consumption modeling.
2/16/2026
VI-CuRL: Stabilizing Verifier-Independent RL Reasoning via Confidence-Guided Variance Reduction
Verifier-independent RL method for LLM reasoning without external verifiers using confidence-guided variance reduction.
2/16/2026
Multi-Head Attention as a Source of Catastrophic Forgetting in MoE Transformers
Analysis of catastrophic forgetting in mixture-of-experts transformers with multi-head attention.
2/16/2026
Power Interpretable Causal ODE Networks: A Unified Model for Explainable Anomaly Detection and Root Cause Analysis in Power Systems
Causal ODE networks for explainable anomaly detection and root cause analysis in power grids.
2/16/2026
RelBench v2: A Large-Scale Benchmark and Repository for Relational Data
RelBench v2 benchmark for relational deep learning on database-like data at scale.
2/16/2026
Coden: Efficient Temporal Graph Neural Networks for Continuous Prediction
Temporal graph neural networks optimized for continuous prediction on dynamic graphs.
2/16/2026
Efficient Personalized Federated PCA with Manifold Optimization for IoT Anomaly Detection
Federated PCA with personalization and manifold optimization for anomaly detection in IoT networks.
2/16/2026
Dual-Granularity Contrastive Reward via Generated Episodic Guidance for Efficient Embodied RL
Dual-granularity contrastive reward framework using generated episodic guidance for sample-efficient embodied reinforcement learning.
2/16/2026
SLA2: Sparse-Linear Attention with Learnable Routing and QAT
SLA2 improves sparse-linear attention for diffusion models with learnable routing and quantization-aware training.
2/16/2026
Trust the uncertain teacher: distilling dark knowledge via calibrated uncertainty
Knowledge distillation via calibrated uncertainty preserves dark knowledge from teachers trained with uncertainty-aware losses.
2/16/2026
Mixture of Predefined Experts: Maximizing Data Usage on Vertical Federated Learning
Split-MoPE addresses sample misalignment in vertical federated learning using mixture of experts specialized for partial data.
2/16/2026
ADEPT: RL-Aligned Agentic Decoding of Emotion via Evidence Probing Tools -- From Consensus Learning to Ambiguity-Driven Emotion Reasoning
ADEPT framework combines speech LLMs with evidence probing tools via RL-aligned agentic decoding for interpretable emotion reasoning.
2/16/2026
Closing the Loop: A Control-Theoretic Framework for Provably Stable Time Series Forecasting with LLMs
Control-theoretic framework addressing stability issues in LLM-based time series forecasting through closed-loop feedback instead of naive autoregressive generation.
2/16/2026
Can Neural Networks Provide Latent Embeddings for Telemetry-Aware Greedy Routing?
Placer applies message passing neural networks to network routing, improving explainability of ML-based telemetry-aware routing decisions.