6/1/2026
VikingMem: A Memory Base Management System for Stateful LLM-based Applications
Memory management system for stateful LLM applications that generalizes memory extraction across use cases beyond fixed context windows.
6/1/2026
SAAS: Self-Aware Reinforcement Learning for Over-Search Mitigation in Agentic Search
Framework enabling AI agents to recognize knowledge boundaries and avoid unnecessary searches, improving efficiency of LLM-based agentic systems.
6/1/2026
OmniMatBench: A Human-Calibrated Multimodal Reasoning Benchmark Across 19 Materials Science Subfields
Benchmark for evaluating multimodal LLMs across 19 materials science subfields, testing reasoning from knowledge to practical application.
6/1/2026
Domain-Specific Data Synthesis for LLMs via Minimal Sufficient Representation Learning
Method for synthesizing domain-specific training data for LLMs using minimal sufficient representation learning to address data acquisition challenges in fine-tuning.
6/1/2026
MIRA: Mid-training Rubric Anchoring for Source-Aware Data Selection
MIRA method for source-aware mid-training data selection balancing pretraining and downstream capability optimization in LLMs.
6/1/2026
Graph Machine Learning in the Era of Large Language Models (LLMs)
Survey of graph machine learning integration with large language models for tasks in social networks and knowledge graphs.
6/1/2026
Breaking Information Cocoons: A Hyperbolic Framework for Balancing Exploration and Exploitation in Recommender Systems
Hyperbolic geometry framework for recommender systems balancing content exploration and exploitation in user preferences.
6/1/2026
Understanding the Fundamental Design Decisions of Retrieval-Augmented Generation Systems
Comprehensive study of fundamental engineering trade-offs in RAG system design including retrieval, generation, and augmentation decisions.
6/1/2026
Cross-Modal Attention Calibration for LVLM Hallucination Mitigation
Cross-modal attention calibration technique reducing hallucinations in large vision-language models during generation.
6/1/2026
Beyond Memorization: Assessing Semantic Generalization in Large Language Models Using Phrasal Constructions
Diagnostic evaluation using Construction Grammar to assess LLM generalization beyond memorization on out-of-domain language.
6/1/2026
PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data Selection
PRISM method for training-free multimodal data selection reducing dataset redundancy in visual instruction tuning.
6/1/2026
Auto-Discovery-Bench: Diagnosing Structured State Tracking in Oracle-Guided Discovery
Auto-Discovery-Bench diagnostic benchmark for evaluating agent state tracking and hypothesis-intervention-feedback cycles.
6/1/2026
EMCEE: Improving Multilingual Capability of LLMs via Bridging Knowledge and Reasoning with Extracted Synthetic Multilingual Context
EMCEE method improving multilingual LLM capabilities by extracting synthetic multilingual context bridging knowledge and reasoning.
6/1/2026
How does Bayesian Sampling help Membership Inference Attacks?
Novel Bayesian sampling approach for membership inference attacks reducing computational overhead of reference model training.
6/1/2026
Unraveling LoRA Interference: Orthogonal Subspaces for Robust Model Merging
Orthogonal subspace approach for merging multiple LoRA-fine-tuned language models without performance degradation.
6/1/2026
Who Gets Credit or Blame? Attributing Accountability in Modern AI Systems
Framework for attributing model behavior to specific training stages (pretraining, fine-tuning, adaptation) in multi-stage AI systems.
6/1/2026
Unlearning's Blind Spots: Over-Unlearning and Prototypical Relearning Attack
Analysis of machine unlearning vulnerabilities including over-unlearning effects and prototypical relearning attacks.
6/1/2026
SHIELD: Secure Hypernetworks for Incremental Expansion Learning Defense
SHIELD framework combining interval bound propagation with hypernetworks for certifiably robust continual learning.
6/1/2026
DISCO: Mitigating Bias in Deep Learning with Conditional Distance Correlation
DISCO framework using conditional distance correlation and causal theory to mitigate dataset bias in deep learning models.
6/1/2026
Organizational Adaptation to Generative AI in Cybersecurity
Qualitative study examining how cybersecurity organizations adapt to generative AI through modified frameworks and hybrid processes.
6/1/2026
PictSure: Pretraining Embeddings Matters for In-Context Learning Image Classifiers
Vision-only in-context learning models for few-shot image classification emphasizing importance of encoder pretraining.
6/1/2026
Joint angle based learning to refine kinematic human pose estimation
Deep learning approach for refining human pose estimation using joint angle learning to reduce keypoint recognition errors.
6/1/2026
Human-Alignment and Calibration of Inference-Time Uncertainty in Large Language Models
Methods for calibrating uncertainty in LLM inference to improve model control and user trust in real-time applications.
6/1/2026
Target-Agnostic Calibration under Distribution Shift with Frequency-Aware Gradient Rectification
Method for improving neural network calibration under distribution shift without requiring target domain access, using frequency-aware gradient rectification.
6/1/2026
Reasoning-Intensive Regression
Introduces reasoning-intensive regression using LLMs to deduce subtle numerical scores from text for rubric-based scoring and reward modeling.
6/1/2026
Human Psychometric Questionnaires Mischaracterize LLM Behavior
Examines whether human psychometric questionnaires reliably characterize LLM behavior by comparing self-report and generation probability methods.
6/1/2026
MedFact: Benchmarking the Fact-Checking Capabilities of Large Language Models on Chinese Medical Texts
Introduces MedFact benchmark with 2,116 expert-annotated Chinese medical texts to evaluate fact-checking capabilities of LLMs in healthcare.
6/1/2026
Towards Atoms of Large Language Models
Proposes Atom Theory to define and identify fundamental representational units in LLMs using non-Euclidean atomic inner product metric.
6/1/2026
Towards Foundation Models for Zero-Shot Time Series Anomaly Detection: Leveraging Synthetic Data and Relative Context Discrepancy
Introduces TimeRCD foundation model for zero-shot time series anomaly detection using synthetic data and relative context discrepancy scoring.
6/1/2026
SAEmnesia: Erasing Concepts in Diffusion Models with Supervised Sparse Autoencoders
SAEmnesia uses supervised sparse autoencoders to erase concepts in diffusion models by enforcing one-to-one concept-neuron mappings.
6/1/2026
Spectral Collapse Drives Loss of Plasticity in Deep Continual Learning
Shows Hessian spectral collapse causes loss of plasticity in continual learning, derives conditions for successful learning of new tasks.
6/1/2026
Dual Mechanisms of Value Expression: Intrinsic vs. Prompted Values in Large Language Models
Analyzes intrinsic vs. prompted value expression mechanisms in LLMs, examining whether they overlap or rely on distinct underlying mechanisms.
6/1/2026
LLMs Lean on Priors, Not Programming Language Semantics
Investigates whether LLMs condition reasoning on formal programming semantics or learned statistical priors through systematic program execution analysis.
6/1/2026
OBCache: Optimal Brain KV Cache Pruning for Efficient Long-Context LLM Inference
OBCache optimally prunes key-value cache in LLMs for efficient long-context inference using true attention impact rather than heuristic rankings.
6/1/2026
PAC-Bayesian Reinforcement Learning Trains Generalizable Policies
Derives PAC-Bayesian generalization bounds for reinforcement learning accounting for Markov dependencies, enabling non-vacuous certificates for modern RL.
6/1/2026
Boundary-Guided Policy Optimization for Memory-efficient RL of Diffusion Large Language Models
Proposes boundary-guided policy optimization for memory-efficient reinforcement learning of diffusion large language models without full likelihood computation.
6/1/2026
CaptionFormer: Unified Segmentation, Tracking, and Captioning for Spatio-Temporal Objects
CaptionFormer jointly performs object detection, tracking, and captioning on video trajectories using unified spatio-temporal approach.
6/1/2026
InfiMed-ORBIT: Aligning LLMs on Open-Ended Complex Tasks via Rubric-Based Incremental Training
InfiMed-ORBIT aligns LLMs on open-ended medical tasks using rubric-based incremental training with reinforcement learning for complex dialogue.
6/1/2026
Scaling Multi-Agent Environment Co-Design with Diffusion Models
Proposes diffusion model-based approach to scale multi-agent environment co-design, enabling joint optimization of agent policies and environment configurations.
6/1/2026
SpectralTrain: A Universal Framework for Hyperspectral Image Classification
Introduces SpectralTrain framework combining curriculum learning and PCA for efficient hyperspectral image classification in remote sensing.
6/1/2026
Mixture of Horizons in Action Chunking
Studies action chunk length trade-offs in vision-language-action robotic models, proposes mixture of horizons approach for improved performance.
6/1/2026
Reasoning-Aware Multimodal Fusion for Hateful Video Detection
Proposes reasoning-aware multimodal fusion framework for detecting hateful content in videos using semantic relationships between modalities.
6/1/2026
Conditional Coverage Diagnostics for Conformal Prediction
Research on evaluating conditional coverage in conformal prediction methods for assessing reliability of predictive systems.
6/1/2026
Bottom-up Policy Optimization: Your Language Model Policy Secretly Contains Internal Policies
Bottom-up policy optimization decomposes LLM policies into internal layer/modular policies; applies RL via Transformer residual streams.
6/1/2026
FEM-Bench: A Structured Scientific Reasoning Benchmark for Evaluating Code-Generating LLMs
FEM-Bench benchmark for evaluating code-generating LLMs on computational mechanics reasoning and scientific model validity.
6/1/2026
Flow Equivariant World Models: Memory for Partially Observed Dynamic Environments
World models incorporating flow equivariance for embodied systems handling partial observability with continuous sensory input.
6/1/2026
Rethinking Multimodal Few-Shot 3D Point Cloud Segmentation: From Fused Refinement to Decoupled Arbitration
Decoupled-experts architecture for few-shot 3D point cloud segmentation resolving plasticity-stability dilemma.
6/1/2026
The Refutability Gap: Challenges in Validating Reasoning by Large Language Models
Critique of LLM reasoning claims arguing they lack falsifiability; identifies methodological pitfalls in AI research validation.
6/1/2026
PASTA: A Scalable Framework for Multi-Policy AI Compliance Evaluation
PASTA framework for scalable multi-policy AI compliance evaluation with model-card format and four integration innovations.
6/1/2026
Performance and Complexity Trade-off Optimization of Speech Models During Training
Method optimizes speech model architecture layer sizes during training balancing performance-complexity trade-offs.