Diagnostic framework for predicting closed-loop performance of world models in model-based RL without explicit validation metrics.
Bayesian reinforcement learning method addressing data scarcity through prior knowledge and belief updates in sequential decision-making.
Medical video benchmark for time-aware clinical AI predictions. Evaluates when models should answer vs defer in real-world deployment.
Decentralized optimization with communication compression for nonsmooth problems. Distributed computing theory, marginal relevance.
Open-vocabulary object detection calibration using frozen VLMs. Vision-language model application with limited novelty.
Physics-informed neural networks and GNNs for RF map construction and multipath propagation. Wireless domain-specific, limited relevance.
Adaptive expert pruning and growing for efficient MoE fine-tuning using LoRA. Parameter-efficient training for large models.
Safe reinforcement learning for UAV navigation with explicit safety mechanisms. Robotics application with limited relevance.
EEG-based cognitive load assessment for online learning using deep learning. Healthcare/education application, minimal AI research depth.
Expander sparse autoencoders for mechanistic interpretability with reduced parameters. ML research on interpretability and efficient dictionaries.
Causal study of AI coding agent adoption on open-source projects, analyzing impact on newcomer participation. Developer tools and OSS ecosystem.
Continuously evolving multimodal benchmark using multi-agent pipeline for VLM evaluation. Developer tools and evaluation frameworks.
Memory-efficient training stack combining parallelism techniques for MoE models. ML research on scalable training infrastructure.
Music decompilation framework recovering executable programs from MIDI via post-training. Niche domain-specific application.
Evaluation of chunking strategies for RAG systems on academic texts using RAGAS framework. Directly relevant to LLM applications and retrieval techniques.
Empirical study of LLM-generated code and comments in real repositories, analyzing prevalence and quality concerns. Developer tools and LLM applications.
Binarization technique for vision-language models reducing memory/latency for deployment. Model optimization for efficient inference.
Reward-free reinforcement learning from video using VLM as progress scorer and GRPO objective. AI agents and novel training approach.
Brain disease diagnosis framework integrating LLM semantics with brain connectivity analysis via hypergraphs. LLM application with healthcare focus.
Population-based evolutionary training for semi-supervised GANs with multi-objective optimization. Limited relevance to core interests.
Image transmission framework using VQ-VAE for latency reduction under spectrum constraints. Not directly relevant to user interests.
Pipeline for adapting Qwen 27B model to perform reasoning in Turkish rather than English, addressing multilingual LLM reasoning.
Dataset of 1,639 K-12 science explanations with human and LLM-generated alternatives for training risk assessment auditors.
CausalSTeward agentic divide-conquer-combine system for causal discovery integrating prior knowledge to identify causal models from high-dimensional data.
PhysMani framework combines physics-principled 3D Gaussian world model with action policy for dynamic object manipulation in embodied AI.
Conditional co-ablation technique reveals transformer self-repair mechanisms where dormant backups activate after primary component ablation.
Analyzes representational geometry showing LLMs become robust to science skepticism through problematic mechanisms rather than genuine understanding.
Novel view video synthesis method from single images using pre-trained video models without task-specific fine-tuning.
Object Aligner provides configurable JSON schema similarity scoring for measuring LLM output structure alignment in tool calling and agentic systems.
Evaluates Vision-Language Model reliability for medical image quality assessment under corruption and bias conditions.
DCASE 2026 Challenge system for audio classification using CLAP audio-text representations with taxonomy-aware hierarchy constraints.
MolSight vision-language model combines molecular LLMs with graph-aware visual understanding for molecular structure and drug discovery tasks.
Controlled study comparing nine lightweight CNN architectures across multiple datasets and hardware to assess efficiency claims.
Proposes load-aware prefill deflection technique to improve disaggregated LLM serving efficiency by balancing prefill and decode GPU pools.
OpenSafeIntent benchmark evaluates whether LLMs calibrate assistance appropriately across benign, dual-use, and malicious intent variants.
SPLIT benchmark evaluates LLM cross-lingual empathy and cultural grounding in emotional-support contexts across English and Ukrainian.
Demonstrates performance evaluation failures in spatiotemporally correlated domains due to data leakage from non-i.i.d. splits.
Proposes prompt coverage adequacy testing framework to guide LLM and autonomous agent testing when prompts replace traditional code.
Graph Neural Network model for EEG-based depression recognition using hyperbolic geometry to capture hierarchical brain network structure.
kNNGuard presents training-free guardrail for LLMs using activation space of off-the-shelf models to detect unsafe/adversarial prompts with minimal labeled data.
Combines Wave Function Collapse procedural generation with evolutionary search to evolve input examples for level generation.
Introduces emotional self-correction mechanism for vision-language models to improve reasoning reliability without post-training or engineered feedback.
Proposes test-time guidance framework for vision-language-action policies using learned critic to guide flow-matching inference without retraining base models.
Presents vLLM-based inference pipeline for unified audio understanding and generation in speech language models with multi-token prediction support.
Develops behavioral monitoring techniques to detect and analyze guardrail activations in LLMs, enabling black-box security testing of production AI systems.
Proposes Adaptive Reparameterized Time (ART) continuous-time control for optimizing timestep allocation in score-based diffusion sampling via actor-critic learning.
Applies deep neural networks and ensemble methods to predict early-stage Alzheimer's disease and identify biomarkers from medical data.
Develops neural graph encoding method for analyzing neural network weight spaces by capturing sequential layer-by-layer inference processes.
Proposes RadiomicNet hybrid architecture integrating handcrafted radiomics features with deep learning for interpretable medical image segmentation.
Reviews AI risk assessment and management methodologies under EU AI Act and other regulatory frameworks, covering identification, analysis, and mitigation approaches.