arXiv paper presenting evolutionary framework using LLMs to discover novel reinforcement learning algorithms by searching over executable update rules.
arXiv paper introducing KGroups, a feature selection algorithm for high-dimensional biological data using max-relevance min-redundancy criteria.
IsoQuant uses quaternion algebra and isoclinic rotations for efficient LLM KV cache compression with hardware-aligned blockwise operations.
FeDMRA addresses federated class-incremental learning with dynamic memory replay for non-IID distributed healthcare data.
HISA improves efficiency of token-level sparse attention mechanisms through hierarchical indexing, reducing O(L²) bottleneck.
Analysis of scaling laws in AI across model families, explaining their predictive power and universal effectiveness in training loss reduction.
Interpretable machine learning framework for detecting low left ventricular ejection fraction from ECG data using predictor-driven approach.
CirrusBench evaluates LLM-based agents in real-world cloud service environments beyond correctness, measuring robustness and efficiency.
Simplex denoising framework for discrete generative modeling using non-Markovian noising scheme, applied to graph generation.
Offline multi-agent reinforcement learning approach using Partial Action Replacement to handle exponential joint action space growth.
ChemCLIP uses contrastive learning to bridge organic and inorganic anticancer compound discovery by enabling knowledge transfer across chemical domains.
Physics-informed impact identification framework combining observational and inductive biases for aerospace composite damage detection.
Position paper arguing explainable AI fundamentally relies on causal inference rather than diverse disconnected XAI methods.
LACE mechanism for continual learning that adaptively expands model capacity during training based on loss signal monitoring.
Information-theoretic analysis of safety verification impossibility for self-improving systems balancing bounded risk with unbounded utility.
Federated learning defense using PPA and game theory minimax approaches against backdoor attacks from malicious clients.
AMIGO benchmark for evaluating agentic vision-language models on long-horizon multi-image grounding tasks through sequential attribute-focused queries.
FL-PBM method for detecting and mitigating backdoor attacks in federated learning during pre-training phases.
PACE system for backpropagation-free test-time adaptation by optimizing normalization layer parameters using covariance matrix approaches.
GPU-accelerated TensorRT inference pipeline for BERT and GPT-2 with mixed-precision optimization achieving 64.4x CPU speedup.
Stepwise credit assignment method for GRPO on flow models that differentiates early composition steps from late detail-refinement steps.
VeoPlace uses vision-language models for chip floorplanning macro placement by leveraging VLM spatial reasoning abilities to complement learning-based approaches.
Theoretical analysis of transfer learning in linear regression and neural networks with closed-form generalization error bounds.
HyperP introduces hypersphere parameterization for language model scaling with improved training stability compared to first-order optimizer approaches.
Analysis of why linear probes and sparse autoencoders fail at compositional generalization under superposition, proposing iterative coding alternatives.
Research on recurrent network training using temporal credit through hidden states without Jacobian propagation, addressing gradient normalization in online adaptation.
SimulCost benchmark evaluates LLM agents on physics simulation tasks with cost-aware metrics, accounting for simulation time and experimental resource usage beyond token costs.
Conversational query rewriting approach for multimodal image retrieval with multi-turn dialogue dataset.
GeoBlock infers optimal block sizes for diffusion language models by analyzing token dependency geometry to enable efficient parallel decoding.
Method for extracting task-relevant contextual representations from 3D scene graphs for robot planning and perception.
Learning-to-rank system for recommending charging nodes in peer-to-peer EV energy trading networks.
Generative theory for synthesizing physiologically consistent multimodal ECG data using quantum-inspired approach.
Benchmark dataset for remote pulse wave detection using event cameras and multimodal physiological sensing.
Physics-informed generative model using Mamba architecture for efficient protein backbone design with linear-time complexity and improved structural fidelity.
FEMBA: bidirectional Mamba state-space model pre-trained on 21k hours EEG with physiologically-aware objectives for microcontroller deployment.
Vision Transformer applied to stress classification from ECG signals using STFT spectrograms for physiological signal analysis.
Enhanced mixture-of-experts architecture using soft nearest neighbor loss to prevent expert collapse and redundant representations.
Cross-lingual evaluation of vision-language models on visual reasoning tasks across Indian languages, revealing performance disparities.
Framework integrating sparse autoencoders with dynamic head pruning in Vision Transformers for interpretable and controllable efficiency.
Deep image clustering using Transformers with distribution information fusion to improve clustering on high-dimensional multimedia data.
MotionGPT3 replaces diffusion with rectified flow objectives for efficient text-driven motion generation with improved convergence.
Evolution strategies warm-start reinforcement learning agents for industrial continuous control using CMA-ES-generated demonstrations.
LogicDiff improves reasoning in masked diffusion language models by prioritizing logical connective tokens during inference-time denoising.
Application of cognitive map learners for hierarchical, compositional robot arm control without task-specific retraining.
Deep learning framework for forecasting black hole plasma accretion dynamics from single telescope images using long-term sequence modeling.
Privacy-preserving inference for spiking neural networks using fully homomorphic encryption to enable encrypted computation.
Language-conditioned multi-game procedural level generation using shared neural representations across different game domains.
Adaptation of Firefly Algorithm metaheuristic for optimization problems with mixed continuous, ordinal, and categorical variables.
Deep learning method for genetic biomarker prediction from pathology images using dictionary-based representations and classifier debiasing.
Benchmark evaluating implicit neural representations for zebrafish brain atlas registration and microscopy reconstruction tasks.