Systematic analysis of targeted instruction selection for LLM fine-tuning, isolating individual component contributions and best practices.
Reduces computational and memory costs in neural network training using randomized, unbiased approximations of vector-jacobian products.
D2-LoRA combines differential and directional low-rank adaptation for parameter-efficient LLM fine-tuning with algebraic mergeability.
Parameter-minimal neural architecture for solving differential equations using Horner-factorized polynomials.
Method to unlock latent capabilities in pretrained transformers through inner loop inference without additional training.
Theoretical framework for meta-learning defining practical universality and distinguishing algorithm-implicit learning capabilities.
Uses graph neural networks for algorithm selection in combinatorial auctions by learning to predict instance-dependent hardness.
Analyzes stability of nonlinear dynamics in gradient descent and SGD beyond quadratic potentials during neural network training.
Extends multi-source Bayesian optimization framework with causality principles for optimization across multiple information sources.
Converts state-tracking tasks from sequence-to-sequence to next-token prediction format for training language models with linear RNNs.
Proposes Interactionless Inverse Reinforcement Learning to decouple alignment artifact learning from policy optimization in AI systems.
Atomix runtime providing transactional semantics for LLM agent tool use with rollback capability and side-effect safety.
Neural network solvers for partial differential equations with formal verification and bounded error guarantees.
Method for comparing clustering algorithms when data contains overlaps and outliers in unsupervised learning.
Proposes curriculum learning approach (Goldilocks RL) to address sparse reward problem in RL for LLM reasoning by tuning task difficulty.
Theoretical analysis of training dynamics in reinforcement learning with verifiable rewards for transformers on compositional reasoning tasks.
Lightweight multimodal summarization framework combining web search with fine-tuned CLIP for semantic image-text alignment and summary generation.
Neural Process-based method for selecting specialized model tools in healthcare AI agents, routing queries to best-performing models per input.
Analysis of conformal prediction under distribution shift, deriving coverage guarantees for pseudo-calibrated methods using domain adaptation tools.
Theoretical analysis showing additive control variates outperform self-normalized inverse propensity scoring for off-policy evaluation in ranking systems.
Unsupervised hypergraph neural network approach addressing performance degradation on heterophilic hypergraphs without requiring labeled data.
Machine unlearning method using forget set gradients for variance-reduced (ε,δ)-unlearning with formal privacy guarantees on data removal.
Online learning approach for multi-objective prediction with adaptive algorithms handling arbitrary distribution shift and multiple simultaneous objectives.
Generative modeling framework for macrocycles incorporating topological guidance to enforce ring-shaped molecular constraints in drug discovery.
Multimodal contrastive learning method using orthogonalization and asymmetric masking to capture modality-specific and synergistic signal interactions.
Geometric deep learning technique extending spectral convolution to orbifolds for supervised learning on non-Euclidean structured data.
New jailbreak attack method (BPJ) that automatically evades classifier-based safeguards in frontier LLMs without requiring white/grey-box access.
Theoretical analysis of discrete diffusion model sampling efficiency with sharp convergence guarantees for KL divergence using tau-leaping samplers.
First scaling law study comparing masked diffusion and uniform-state discrete diffusion language models, showing masked diffusion performance characteristics.
Novel approach to diffusion models using canonicalization instead of architectural constraints for molecular graph generation with symmetry invariance.
Benchmark study (PAPerBench) investigating how context length affects privacy leakage and personalization effectiveness in large language models at scale.
Study of geometric structures in LLM representations, showing how language statistics symmetries shape emergent geometric patterns like circular month organization and linear manifolds.
Research on source bias in dense retrieval systems, where LLM-generated text is preferentially ranked over human text due to lower perplexity in neural information retrieval.
LLM-enabled reinforcement learning framework for optimizing wireless networks with distributed intelligence.
Agentic AI system for commercial insurance underwriting with adversarial self-critique and human-in-the-loop reliability.
SSLogic: agentic meta-synthesis approach scaling logical reasoning tasks with verifiable training signals for RLVR.
Theoretical framework modeling intelligence as trajectory-level Pareto optimization over time.
Quantum reinforcement learning approach for optimizing stacked intelligent metasurfaces in wireless security.
Boltz foundation model for atom-level representation learning in molecular property prediction tasks.
Framework for instilling ethical competence into AI decision-making models with performance metrics.
XAI-enhanced deep learning intrusion detection framework for cybersecurity with interpretability.
TemporalBench evaluates LLM-based agents on temporal reasoning with contextual and event-informed time series tasks.
Explanatory interactive learning to mitigate gender classification bias through user-guided model training.
Benchmarking deep learning models (GRU, TCN, Transformer, TSMixer) for anomaly detection across cloud telemetry.
Metrics and decomposition methods for evaluating HRRP generative models in radar applications.
Conditional generative models for radar high-resolution range profiles using maritime dataset.
Analysis of spectral collapse in diffusion inversion for unpaired image-to-image translation.
Fine-tuned vision-language model for automated artistic creativity scoring with explanatory feedback.
DECKBench evaluates multi-agent frameworks for academic slide generation and editing tasks.
Automated linguistic feature extraction for detecting jailbreak attempts in clinical training LLMs.