2/26/2026
SciTS: Scientific Time Series Understanding and Generation with LLMs
Framework for scientific time series understanding and generation using LLMs, addressing gaps in multimodal LLM handling of temporal numerical data.
2/26/2026
ImpMIA: Leveraging Implicit Bias for Membership Inference Attack
Membership inference attack method using implicit bias to determine which data samples trained a model, focusing on privacy implications.
2/26/2026
Rethinking Evaluation in the Era of Time Series Foundation Models: (Un)known Information Leakage Challenges
Research on information leakage in Time Series Foundation Models evaluation, addressing test set integrity challenges similar to LLM benchmarking issues.
2/26/2026
MNO: Multiscale Neural Operator for 3D Computational Fluid Dynamics
Neural operator architecture for solving PDEs on 3D unstructured domains using multiscale decomposition for computational fluid dynamics.
2/26/2026
SPACeR: Self-Play Anchoring with Centralized Reference Models
SPACeR: Self-play framework with centralized reference models for training human-like autonomous vehicle policies.
2/26/2026
SERL: Self-Examining Reinforcement Learning on Open-Domain
SERL: Self-examining RL approach for LLMs on open-domain tasks using intrinsic reward signals without external feedback.
2/26/2026
EARL: Entropy-Aware RL Alignment of LLMs for Reliable RTL Code Generation
EARL: RL alignment framework with entropy-aware rewards for improving LLM-generated RTL code reliability and correctness.
2/26/2026
NTK-Guided Implicit Neural Teaching
NTK-Guided acceleration of implicit neural representations for high-resolution signal reconstruction tasks.
2/26/2026
Stabilizing Off-Policy Training for Long-Horizon LLM Agent via Turn-Level Importance Sampling and Clipping-Triggered Normalization
Stabilizes off-policy RL training for multi-turn LLM agents using turn-level importance sampling and clipping-triggered normalization.
2/26/2026
BRIDGE: Building Representations In Domain Guided Program Synthesis
BRIDGE: Structured prompting framework for LLM-based program synthesis with formal verification in Lean4 across multiple artifacts.
2/26/2026
Efficient Cross-Domain Offline Reinforcement Learning with Dynamics- and Value-Aligned Data Filtering
Cross-domain offline RL with dynamics and value-aligned data filtering for transferring knowledge between source and target environments.
2/26/2026
NRGPT: An Energy-based Alternative for GPT
NRGPT: Alternative to GPT architecture using energy-based modeling paradigm for language inference.
2/26/2026
WebGym: Scaling Training Environments for Visual Web Agents with Realistic Tasks
WebGym: Large-scale open-source environment with 300K realistic web tasks for training visual web agents via reinforcement learning.
2/26/2026
Stagewise Reinforcement Learning and the Geometry of the Regret Landscape
Extends singular learning theory to reinforcement learning, characterizing policy posterior concentration via regret landscape geometry.
2/26/2026
Orthogonalized Policy Optimization:Policy Optimization as Orthogonal Projection in Hilbert Space
OPO: Principled framework for LLM alignment reformulated as orthogonal projection in Hilbert space for policy optimization.
2/26/2026
Principled Latent Diffusion for Graphs via Laplacian Autoencoders
Latent diffusion approach using Laplacian autoencoders for efficient graph generation with reduced computational complexity.
2/26/2026
OGD4All: A Framework for Accessible Interaction with Geospatial Open Government Data Based on Large Language Models
OGD4All: LLM-based framework with agentic reasoning for accessible interaction with geospatial open government data via code generation.
2/26/2026
DCoPilot: Generative AI-Empowered Policy Adaptation for Dynamic Data Center Operations
DCoPilot: Generative AI system using reinforcement learning for adaptive policy generation in dynamic data center operations.
2/26/2026
QTALE: Quantization-Robust Token-Adaptive Layer Execution for LLMs
QTALE: Framework combining token-adaptive layer execution with quantization for efficient LLM deployment, addressing integration challenges.
2/26/2026
Optimizer choice matters for the emergence of Neural Collapse
Analyzes role of optimizer choice in Neural Collapse emergence during deep neural network training, challenging assumption that NC is universal across optimization methods.
2/26/2026
Quad Length Codes for Lossless Compression of e4m3
Proposes Quad Length Codes for lossless compression of e4m3 format to reduce network bandwidth bottlenecks in LLM training and serving through faster decoding than Huffman codes.
2/26/2026
Capabilities Ain't All You Need: Measuring Propensities in AI
Proposes measuring AI propensities (behavioral tendencies) alongside capabilities using Item Response Theory extensions for evaluation.
2/26/2026
Decoding as Optimisation on the Probability Simplex: From Top-K to Top-P (Nucleus) to Best-of-K Samplers
Unifies decoding strategies (greedy, top-k, nucleus, best-of-k) as principled optimization on probability simplex with regularization.
2/26/2026
Training-Free Generative Modeling via Kernelized Stochastic Interpolants
Kernel-based generative modeling within stochastic interpolant framework replacing neural network training with linear systems.
2/26/2026
MultiModalPFN: Extending Prior-Data Fitted Networks for Multimodal Tabular Learning
MultiModalPFN extends TabPFN foundation model to handle heterogeneous tabular and non-tabular modalities like images and text.
2/26/2026
QuantVLA: Scale-Calibrated Post-Training Quantization for Vision-Language-Action Models
QuantVLA: post-training quantization framework for vision-language-action models to reduce compute and memory for embodied agents.
2/26/2026
Does Order Matter : Connecting The Law of Robustness to Robust Generalization
Theoretical analysis connecting law of robustness to robust generalization in overparameterized models.
2/26/2026
SELAUR: Self Evolving LLM Agent via Uncertainty-aware Rewards
Self-evolving LLM agent framework using uncertainty-aware rewards to improve multi-step decision-making and credit assignment.
2/26/2026
Hardness of Maximum Likelihood Learning of DPPs
Analyzes computational hardness of maximum likelihood learning for Determinantal Point Processes used in data selection.
2/26/2026
Private Blind Model Averaging - Distributed, Non-interactive, and Convergent
Distributed differentially private learning enabling multiple users to jointly train models without centralizing training data.
2/26/2026
The Dark Side of ChatGPT: Legal and Ethical Challenges from Stochastic Parrots and Hallucination
Examines legal and ethical challenges from ChatGPT and large language models including hallucination and stochastic parrots.
2/26/2026
Incentive-Aware Synthetic Control: Accurate Counterfactual Estimation via Incentivized Exploration
Econometric analysis of synthetic control methods for treatment effect estimation when units select their own interventions.
2/26/2026
Multi-agent deep reinforcement learning with centralized training and decentralized execution for transportation infrastructure management
Multi-agent deep reinforcement learning with centralized training for transportation infrastructure lifecycle management and maintenance optimization.
2/26/2026
MPC of Uncertain Nonlinear Systems with Meta-Learning for Fast Adaptation of Neural Predictive Models
Uses neural state-space models with meta-learning for model predictive control of nonlinear systems with fast adaptation.
2/26/2026
Temporal Knowledge-Graph Memory in a Partially Observable Environment
Introduces Room Environment v3 benchmark for agents with graph-structured memory in partially observable environments using knowledge graphs.
2/26/2026
Renaissance: Investigating the Pretraining of Vision-Language Encoders
Investigates best practices for pretraining vision-language encoders and provides programming tools for VL research.
2/26/2026
A Distributional Treatment of Real2Sim2Real for Object-Centric Agent Adaptation in Vision-Driven Deformable Linear Object Manipulation
Real2Sim2Real framework for deformable linear object manipulation using likelihood-free inference and visual perception for robotic agents.
2/26/2026
Measuring AI Ability to Complete Long Software Tasks
Proposes metric measuring AI ability to complete long software tasks by comparing model performance to human domain expert completion time.
2/26/2026
High-Fidelity And Complex Test Data Generation For Google SQL Code Generation Services
Test data generation method for SQL code generation services using high-fidelity synthetic data to model complex data structures and semantic relationships.
2/26/2026
Discrete Optimal Transport and Voice Conversion
kDOT: discrete optimal transport framework for voice conversion using barycentric projection in pretrained speech embedding space instead of averaging strategies.
2/26/2026
BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMs
BARREL identifies pathological reasoning patterns in Large Reasoning Models and improves factual reliability. Enables models to admit ignorance instead of confident false answers.
2/26/2026
Knowledge Fusion of Large Language Models Via Modular SkillPacks
Knowledge fusion method for LLMs via modular SkillPacks. Enables efficient cross-capability transfer for multi-task integration, compression, and continual learning.
2/26/2026
Hyperbolic recurrent neural network as the first type of non-Euclidean neural quantum state ansatz
First non-Euclidean neural quantum state ansatz using hyperbolic GRU for Variational Monte Carlo approximation of quantum many-body ground states.
2/26/2026
Highly Efficient and Effective LLMs with Multi-Boolean Architectures
Novel multi-Boolean architecture for weight-binarized LLMs. Framework with multi-kernel Boolean parameters reduces complexity without severe post-training performance loss.
2/26/2026
Spurious Rewards: Rethinking Training Signals in RLVR
Shows RLVR with GRPO can improve LLM mathematical reasoning using spurious rewards with little/no correlation to correct answers, challenging reward signal assumptions.
2/26/2026
Blameless Users in a Clean Room: Defining Copyright Protection for Generative Models
Establishes foundations for provable copyright protection in generative models. Revisits near access-freeness and defines conditions for copyright guarantee.
2/26/2026
A Comprehensive Benchmark for Electrocardiogram Time-Series
Comprehensive benchmark for ECG time-series data addressing unique characteristics and specialized downstream applications of bioelectrical signals.
2/26/2026
CASCADE: LLM-Powered JavaScript Deobfuscator at Google
CASCADE: hybrid LLM-powered JavaScript deobfuscator at Google combining Gemini coding capabilities with compiler IR transformations for code comprehension.
2/26/2026
Position: Beyond Sensitive Attributes, ML Fairness Should Quantify Structural Injustice via Social Determinants
Position paper arguing ML fairness research should quantify structural injustice via social determinants rather than focusing only on sensitive attributes.
2/26/2026
Do LLMs Adhere to Label Definitions? Examining Their Receptivity to External Label Definitions
Controlled experiments examining whether LLMs incorporate external label definitions or rely on parametric knowledge. Tests expert-curated, LLM-generated, and perturbed definitions.