3/20/2026
CRAFT: Aligning Diffusion Models with Fine-Tuning Is Easier Than You Think
CRAFT method for aligning diffusion models through fine-tuning, addressing limitations of SFT and DPO-style preference optimization approaches.
3/20/2026
Hypothesis-Conditioned Query Rewriting for Decision-Useful Retrieval
Hypothesis-Conditioned Query Rewriting improves RAG systems by rewriting queries to prioritize decision-relevant evidence over topical relevance.
3/20/2026
Towards Verifiable AI with Lightweight Cryptographic Proofs of Inference
Lightweight cryptographic framework for verifiable AI inference enabling clients to verify model outputs without rerunning computation.
3/20/2026
SEM: Sparse Embedding Modulation for Post-Hoc Debiasing of Vision-Language Models
SEM method for post-hoc debiasing of CLIP via sparse embedding modulation to remove social and spurious biases.
3/20/2026
Fast and Interpretable Autoregressive Estimation with Neural Network Backpropagation
Neural network approach to autoregressive time series estimation using backpropagation while preserving interpretability.
3/20/2026
SAVeS: Steering Safety Judgments in Vision-Language Models via Semantic Cues
SAVeS framework steers safety judgments in Vision-Language Models through semantic cues and textual/visual interventions.
3/20/2026
FedTrident: Resilient Road Condition Classification Against Poisoning Attacks in Federated Learning
FedTrident defends federated learning-based road classification against poisoning attacks from malicious participants.
3/20/2026
How Uncertainty Estimation Scales with Sampling in Reasoning Models
Studies how uncertainty estimation scales with sampling in reasoning language models using self-consistency and verbalized confidence.
3/20/2026
D5P4: Partition Determinantal Point Process for Diversity in Parallel Discrete Diffusion Decoding
D5P4 framework applies determinantal point processes to discrete diffusion decoding for diverse parallel text generation.
3/20/2026
Optimal Splitting of Language Models from Mixtures to Specialized Domains
Method for splitting pretrained language models into specialized domain-specific models using continued pretraining strategies.
3/20/2026
Meanings and Measurements: Multi-Agent Probabilistic Grounding for Vision-Language Navigation
Multi-agent framework for grounding vision-language navigation using probabilistic reasoning about spatial relations and metric constraints.
3/20/2026
Do VLMs Need Vision Transformers? Evaluating State Space Models as Vision Encoders
Evaluates State Space Models as vision encoders for Vision-Language Models, comparing SSM backbones to transformer-based alternatives.
3/20/2026
DreamPartGen: Semantically Grounded Part-Level 3D Generation via Collaborative Latent Denoising
DreamPartGen generates semantically grounded 3D objects with part-level decomposition using text-to-3D diffusion methods.
3/20/2026
DriveTok: 3D Driving Scene Tokenization for Unified Multi-View Reconstruction and Understanding
DriveTok proposes efficient 3D tokenization for multi-view driving scenes to improve autonomous driving systems and world models.
3/20/2026
Nemotron-Cascade 2: Post-Training LLMs with Cascade RL and Multi-Domain On-Policy Distillation
Nemotron-Cascade 2: 30B open-weight MoE LLM with strong reasoning and agentic capabilities, achieving IMO Gold Medal performance.
3/20/2026
NavTrust: Benchmarking Trustworthiness for Embodied Navigation
NavTrust benchmark evaluates trustworthiness of embodied navigation agents under real-world corruptions in Vision-Language Navigation and Object-Goal Navigation tasks.
3/20/2026
Improved Learning Rates for Stochastic Optimization
Establishes improved learning rates for stochastic gradient descent and Nesterov accelerated gradient with generalization performance guarantees.
3/20/2026
Rule Extraction in Machine Learning: Chat Incremental Pattern Constructor
Chat Incremental Pattern Constructor extracts ordered token-transition rules from text for interpretable machine learning rule extraction.
3/20/2026
Inverse classification with logistic and softmax classifiers: efficient optimization
Optimization methods for inverse classification problems including counterfactual explanations and adversarial examples using logistic and softmax classifiers.
3/20/2026
CADGL: Context-Aware Deep Graph Learning for Predicting Drug-Drug Interactions
CADGL uses context-aware deep graph learning for predicting drug-drug interactions with improved generalization and robustness.
3/20/2026
$\mu$LO: Compute-Efficient Meta-Generalization of Learned Optimizers
μLO derives Maximal Update Parametrization for learned optimizers to improve meta-generalization across network widths and unseen tasks.
3/20/2026
Modeling Inverse Ellipsometry Problem via Flow Matching with a Large-Scale Dataset
Flow matching approach with large-scale synthetic dataset for solving inverse ellipsometry problem of reconstructing optical film properties.
3/20/2026
ODE-Constrained Generative Modeling of Cardiac Dynamics for 12-Lead ECG Synthesis
ODE-constrained generative model for synthesizing realistic 12-lead ECG training data to address scarcity of labeled medical recordings.
3/20/2026
Cliqueformer: Model-Based Optimization with Structured Transformers
Cliqueformer uses structured transformers for model-based optimization in design problems like protein engineering via offline learning.
3/20/2026
Vector Optimization with Gaussian Process Bandits
VOGP algorithm using Gaussian process bandits for black-box vector optimization with incomplete order relations and Pareto optimality guarantees.
3/20/2026
Linearly Separable Features in Shallow Nonlinear Networks: Width Scales Polynomially with Intrinsic Data Dimension
Theoretical analysis showing shallow nonlinear networks learn linearly separable features with polynomial width scaling relative to data dimension.
3/20/2026
Unlocking Full Efficiency of Token Filtering in Large Language Model Training
Methods to achieve real-world efficiency gains from token filtering in LLM training through improved sparsity and adaptive filtering strategies.
3/20/2026
This looks like what? Challenges and Future Research Directions for Part-Prototype Models
Survey of Part-Prototype Models for explainable AI, examining interpretability mechanisms and competitive limitations versus alternative approaches.
3/20/2026
Integrating Weather Station Data and Radar for Precipitation Nowcasting: SmaAt-fUsion and SmaAt-Krige-GNet
Two neural architectures for precipitation nowcasting integrating weather station data and radar measurements for improved forecast skill.
3/20/2026
OPUS-VFL: Incentivizing Optimal Privacy-Utility Tradeoffs in Vertical Federated Learning
OPUS-VFL addresses privacy-utility tradeoffs and incentive mechanisms in Vertical Federated Learning with heterogeneous client resources.
3/20/2026
Causal Intervention Framework for Variational Auto Encoder Mechanistic Interpretability
Causal intervention framework for interpreting Variational Autoencoders mechanistically, addressing interpretability of generative models.
3/20/2026
Modality Equilibrium Matters: Minor-Modality-Aware Adaptive Alternating for Cross-Modal Memory Enhancement
Shapley Value-based alternating training framework for multimodal fusion that balances dominant and minor modalities.
3/20/2026
Size-adaptive Hypothesis Testing for Fairness
Statistical framework for fairness testing in algorithmic systems that accounts for sampling error and handles intersectional demographic analysis.
3/20/2026
On the Surprising Effectiveness of a Single Global Merging in Decentralized Learning
Analysis of communication scheduling in decentralized learning showing benefits of concentrating synchronization in later training stages.
3/20/2026
GeoReg: Weight-Constrained Few-Shot Regression for Socio-Economic Estimation using LLM
GeoReg uses LLMs with satellite imagery and geospatial data for socio-economic indicator estimation in data-scarce regions via few-shot regression.
3/20/2026
Online Convex Optimization with Heavy Tails: Old Algorithms, New Regrets, and Applications
Research on Online Convex Optimization algorithms for heavy-tailed gradient distributions, extending beyond finite variance assumptions.
3/20/2026
Pi-transformer: A prior-informed dual-attention model for multivariate time-series anomaly detection
Transformer architecture with dual attention for multivariate time-series anomaly detection using temporal invariants.
3/20/2026
Investigating Faithfulness in Large Audio Language Models
Framework evaluating faithfulness of chain-of-thought reasoning in large audio language models for multimodal tasks.
3/20/2026
OT-MeanFlow3D: Bridging Optimal Transport and Meanflow for Efficient 3D Point Cloud Generation
Flow-matching models for 3D point cloud generation using optimal transport and meanflow for single-step inference acceleration.
3/20/2026
Splines-Based Feature Importance in Kolmogorov-Arnold Networks: A Framework for Supervised Tabular Data Dimensionality Reduction
KAN-based feature selection framework for tabular data via spline-based importance scoring. Specialized ML technique.
3/20/2026
Support Basis: Fast Attention Beyond Bounded Entries
Sub-quadratic attention algorithm removing bounded-entry restrictions for LLM inference speedup. Foundational LLM efficiency research.
3/20/2026
Activation Quantization of Vision Encoders Needs Prefixing Registers
Quantization technique for vision encoders using prefix registers to handle outliers. Optimization research for multimodal models.
3/20/2026
Image2Gcode: Image-to-G-code Generation for Additive Manufacturing Using Diffusion-Transformer Model
Diffusion-Transformer model converting images directly to G-code for 3D printing. Applied ML, domain-specific.
3/20/2026
Heads collapse, features stay: Why Replay needs big buffers
Continual learning research on replay buffer size impact on feature retention vs. classifier forgetting. Specialized ML theory.
3/20/2026
Weights to Code: Extracting Interpretable Algorithms from the Discrete Transformer
Algorithm extraction from Discrete Transformers via symbolic program synthesis. Addresses representation entanglement in interpretability.
3/20/2026
Mechanism Shift During Post-training from Autoregressive to Masked Diffusion Language Models
Research analyzing mechanistic changes when post-training autoregressive models into masked diffusion models. Studies model internals via circuit analysis.
3/20/2026
A Unified Generalization Framework for Model Merging: Trade-offs, Non-Linearity, and Scaling Laws
Unified theoretical framework for model merging explaining effectiveness across heterogeneous fine-tuning hyperparameters with scaling laws.
3/20/2026
Mixed-Precision Training and Compilation for RRAM-based Computing-in-Memory Accelerators
Mixed-precision training and compilation techniques for RRAM-based computing-in-memory ML accelerators with low bit-width constraints.
3/20/2026
Krause Synchronization Transformers
Krause Attention: principled attention mechanism addressing representation collapse and attention sink issues in transformers.
3/20/2026
TS-Haystack: A Multi-Scale Retrieval Benchmark for Time Series Language Models
Multi-scale retrieval benchmark for time series language models addressing long-context temporal localization under computational constraints.