3/23/2026
Improving Automatic Summarization of Radiology Reports through Mid-Training of Large Language Models
Proposes mid-training adaptation strategy for LLMs to improve automatic summarization of radiology reports in clinical domain.
3/23/2026
From Flat to Structural: Enhancing Automated Short Answer Grading with GraphRAG
Enhances automated short answer grading using GraphRAG to improve LLM reasoning through structured knowledge representation instead of flat vector retrieval.
3/23/2026
HypeLoRA: Hyper-Network-Generated LoRA Adapters for Calibrated Language Model Fine-Tuning
Proposes HypeLoRA, a hyper-network framework for parameter-efficient fine-tuning of Transformers, addressing miscalibration in language models via LoRA adaptation.
3/23/2026
From Feature-Based Models to Generative AI: Validity Evidence for Constructed Response Scoring
Evaluates generative AI and LLMs for scoring constructed responses in high-stakes testing, comparing feature-based models to neural approaches.
3/23/2026
URAG: A Benchmark for Uncertainty Quantification in Retrieval-Augmented Large Language Models
URAG: benchmark for evaluating uncertainty quantification in retrieval-augmented generation systems across multiple domains.
3/23/2026
Framing Effects in Independent-Agent Large Language Models: A Cross-Family Behavioral Analysis
Analysis of how prompt framing biases LLM decisions in threshold voting tasks with individual-group interest conflicts.
3/23/2026
CDEoH: Category-Driven Automatic Algorithm Design With Large Language Models
CDEoH: LLM-driven automated algorithm design with category-driven diversity to improve evolutionary stability and convergence.
3/23/2026
Generalized Stock Price Prediction for Multiple Stocks Combined with News Fusion
Stock price prediction approach integrating LLMs with financial news using stock name embeddings for relevance filtering.
3/23/2026
Speculating Experts Accelerates Inference for Mixture-of-Experts
Expert prefetching scheme for accelerating MoE model inference by predicting and preloading expert weights in memory-constrained settings.
3/23/2026
Automatic Analysis of Collaboration Through Human Conversational Data Resources: A Review
Review of automatic collaboration analysis methods using task-oriented conversational data for studying human cooperation.
3/23/2026
LLM-MRD: LLM-Guided Multi-View Reasoning Distillation for Fake News Detection
LLM-MRD: distillation approach for multimodal fake news detection combining LLM reasoning with efficient student models.
3/23/2026
Maximizing mutual information between user-contexts and responses improve LLM personalization with no additional data
Self-improvement framework for LLMs using mutual information maximization between user context and responses without additional labeled data.
3/23/2026
Parameter-Efficient Token Embedding Editing for Clinical Class-Level Unlearning
STEU: parameter-efficient method for unlearning sensitive information from clinical language models while preserving utility.
3/23/2026
Agreement Between Large Language Models, Human Reviewers, and Authors in Evaluating STROBE Checklists for Observational Studies in Rheumatology
Study comparing LLM, human, and author assessments of STROBE reporting checklists in observational rheumatology studies.
3/23/2026
VERDICT: Verifiable Evolving Reasoning with Directive-Informed Collegial Teams for Legal Judgment Prediction
VERDICT: multi-agent LLM system for interpretable legal judgment prediction with verifiable reasoning across law articles and penalties.
3/23/2026
MemReward: Graph-Based Experience Memory for LLM Reward Prediction with Limited Labels
MemReward: graph-based experience memory system for LLM reward prediction with limited labeled data using prior experience.
3/23/2026
LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels
LeWorldModel: stable joint-embedding predictive architecture for learning world models from raw pixels with minimal loss terms.
3/23/2026
Memory-Driven Role-Playing: Evaluation and Enhancement of Persona Knowledge Utilization in LLMs
Memory-driven role-playing paradigm for LLMs using internal memory stores to maintain consistent personas during long dialogues.
3/23/2026
Ternary Gamma Semirings: From Neural Implementation to Categorical Foundations
Theoretical framework connecting neural networks with ternary gamma semiring algebra, demonstrating improved compositional generalization.
3/23/2026
Prompt-tuning with Attribute Guidance for Low-resource Entity Matching
Low-resource entity matching method using prompt-tuning with attribute guidance, reducing need for large labeled datasets.
3/23/2026
Goedel-Code-Prover: Hierarchical Proof Search for Open State-of-the-Art Code Verification
Hierarchical proof search framework using LLMs for automated code verification in Lean 4, decomposing complex verification goals into subgoals.
3/23/2026
PAI: Fast, Accurate, and Full Benchmark Performance Projection with AI
AI-based hardware performance projection system for SoC benchmarking as faster alternative to cycle-accurate simulators.
3/23/2026
POET: Power-Oriented Evolutionary Tuning for LLM-Based RTL PPA Optimization
Framework using LLMs with evolutionary tuning for RTL code optimization prioritizing power reduction while ensuring functional correctness.
3/23/2026
Do Post-Training Algorithms Actually Differ? A Controlled Study Across Model Scales Uncovers Scale-Dependent Ranking Inversions
Unified framework evaluating 51 post-training alignment algorithms across model scales, revealing scale-dependent performance ranking inversions.
3/23/2026
Diffusion-Guided Semantic Consistency for Multimodal Heterogeneity
Federated learning approach using diffusion guidance to handle non-IID multimodal data and semantic discrepancies across clients.
3/23/2026
Spectral Tempering for Embedding Compression in Dense Passage Retrieval
Method for compressing embeddings in dense retrieval systems using spectral tempering to balance variance preservation and noise reduction.
3/23/2026
Scalable Prompt Routing via Fine-Grained Latent Task Discovery
Framework for dynamically routing queries to optimal LLMs from large model pools using fine-grained latent task discovery without manual taxonomies.
3/23/2026
Investigating In-Context Privacy Learning by Integrating User-Facing Privacy Tools into Conversational Agents
Study on teaching users privacy protection through integrated tools within conversational agents via in-context experiential learning.
3/23/2026
The Autonomy Tax: Defense Training Breaks LLM Agents
Analysis of capability-alignment paradox where defense training against prompt injection attacks degrades LLM agent performance on autonomous tool use tasks.
3/23/2026
Is Evaluation Awareness Just Format Sensitivity? Limitations of Probe-Based Evidence under Controlled Prompt Structure
Research investigating whether linear probes detect genuine evaluation awareness in LLMs or merely surface-level format sensitivity in benchmarks.
3/23/2026
Vocabulary shapes cross-lingual variation of word-order learnability in language models
Study on how vocabulary and word order affect learnability in transformer language models across different languages using synthetic variants.
3/23/2026
LoFi: Location-Aware Fine-Grained Representation Learning for Chest X-ray
LoFi: fine-grained location-aware representation learning for chest X-ray retrieval and phrase grounding without region-level labels.
3/23/2026
TrustFlow: Topic-Aware Vector Reputation Propagation for Multi-Agent Ecosystems
TrustFlow: reputation propagation algorithm for multi-agent ecosystems using topic-aware vector reputation instead of scalar scores.
3/23/2026
Global Convergence of Multiplicative Updates for the Matrix Mechanism: A Collaborative Proof with Gemini 3
Analysis of multiplicative update fixed-point iteration convergence for nuclear norm optimization in private machine learning.
3/23/2026
A Framework for Formalizing LLM Agent Security
Framework formalizing security definitions for LLM agents considering contextual action legitimacy and instruction origins.
3/23/2026
Adaptive Layerwise Perturbation: Unifying Off-Policy Corrections for LLM RL
Adaptive layerwise perturbation method addressing off-policy problems in LLM reinforcement learning training stability.
3/23/2026
FedAgain: A Trust-Based and Robust Federated Learning Strategy for an Automated Kidney Stone Identification in Ureteroscopy
FedAgain: trust-based federated learning strategy for robust kidney stone identification from endoscopic images across diverse devices.
3/23/2026
Gastric-X: A Multimodal Multi-Phase Benchmark Dataset for Advancing Vision-Language Models in Gastric Cancer Analysis
Gastric-X: multimodal benchmark dataset for training vision-language models on gastric cancer analysis with clinical workflow alignment.
3/23/2026
Inducing Sustained Creativity and Diversity in Large Language Models
Methods to induce sustained creativity and diversity in LLM outputs for exploratory search tasks requiring iterative discovery.
3/23/2026
dinov3.seg: Open-Vocabulary Semantic Segmentation with DINOv3
dinov3.seg implements open-vocabulary semantic segmentation using vision-language models for pixel-level classification of unseen classes.
3/23/2026
FDARxBench: Benchmarking Regulatory and Clinical Reasoning on FDA Generic Drug Assessment
FDARxBench: benchmark for evaluating LLM document-grounded question-answering on FDA drug labels for regulatory assessment.
3/23/2026
Subspace Kernel Learning on Tensor Sequences
Novel kernel framework (UKTL) for learning from higher-order tensor data with efficient similarity measures across tensor modes.
3/23/2026
Optimal Scalar Quantization for Matrix Multiplication: Closed-Form Density and Phase Transition
Research on scalar quantization methods for matrix multiplication to minimize mean-squared error, with applications to efficient computation.
3/23/2026
PFM-VEPAR: Prompting Foundation Models for RGB-Event Camera based Pedestrian Attribute Recognition
PFM-VEPAR uses prompted foundation models for pedestrian attribute recognition combining RGB and event camera data with minimal computational overhead.
3/23/2026
Evolving Embodied Intelligence: Graph Neural Network--Driven Co-Design of Morphology and Control in Soft Robotics
Uses graph neural networks to co-design soft robot morphology and control policies, solving the challenge of simultaneous optimization through morphology-aware learning.
3/23/2026
Skilled AI Agents for Embedded and IoT Systems Development
Skilled AI Agents addresses hardware-in-the-loop challenges for LLM-based agents in embedded and IoT systems development with timing and peripheral constraints.
3/23/2026
Physics-Informed Neural Network with Adaptive Clustering Learning Mechanism for Information Popularity Prediction
Physics-informed neural networks with adaptive clustering predict information cascade popularity using graph convolution and recurrent architectures.
3/23/2026
CAF-Score: Calibrating CLAP with LALMs for Reference-free Audio Captioning Evaluation
CAF-Score proposes a reference-free metric for evaluating audio captioning by combining CLAP embeddings with large audio-language models.
3/23/2026
DeepStock: Reinforcement Learning with Policy Regularizations for Inventory Management
DeepStock applies policy regularizations from classical inventory theory to deep reinforcement learning for inventory management, improving hyperparameter sensitivity.
3/23/2026
MetaCues: Enabling Critical Engagement with Generative AI for Information Seeking and Sensemaking
MetaCues is an interactive tool using generative AI to promote critical engagement and metacognitive thinking during information seeking tasks.