Search and filter through extracted claims from AI researchers.
Showing 1781-1800 of 2685 claims of type "fact"
Dual-path diffusion models with global representation modulation and local structure refinement can improve infrared image super-resolution consistency while preserving generative capacity
Explicitly encoding causal design principles and using synthetic data generation with known ground-truth effects can make silent failures in automated research systems visible
GKR-HND protocol can verify Transformer inference without requiring clients to replay dense matrix computations, assuming honest verifier and prover-worker non-collusion
Knowledge retrieval and reasoning is the primary bottleneck in Knowledge-Intensive Visual Question Answering, not visual grounding or object identification
Final success accuracy alone is insufficient for evaluating conversational LLM clarification behavior, as models with similar accuracy can differ substantially in efficiency and robustness
Trunk-topology-based frameworks using Delaunay triangulation can solve loop closure detection and global registration in GNSS-denied forest environments despite sparse, noisy LiDAR data
Chest-height pedestrian-view datasets combined with staged target-domain adaptation can enable effective safety-oriented navigation for blind and visually impaired pedestrians
Synthetic augmentation generally improves semantic segmentation accuracy for sidewalk navigation models
Risk Graph Neural Networks substantially improve predictive calibration for heat mortality estimation while preserving interpretable risk curve outputs, maintaining lower point-errors and near-nominal uncertainty coverage during 2022 heatwave
Causal-AgentIR organizes restoration experience as causal graphs connecting degradation patterns, regions, tools, actions, quality changes, and preferences, enabling better knowledge organization than isolated textual records
Quantum state preparation is challenging because search space grows exponentially with number of qubits, making identification of optimal circuit non-trivial
Deep reinforcement learning agent based on proximal policy optimization can achieve quantum state preparation approximation errors of 10^-14 for 2-5 qubit systems
Relative Value Learning achieves competitive performance on Atari benchmark compared to standard PPO, demonstrating that relative value estimation is effective alternative to absolute critics
Replacing object-level visual tokens with compact text proxies can convey the same content in far fewer tokens
GlucoTune framework enables reproducible experimentation with blood glucose time-series data through configurable YAML-based preprocessing pipelines
Common deletion baselines in offline RL have environment-dependent privacy-utility behavior, with retraining and fine-tuning often providing stronger retained-utility than uniform GA+Refit
Evaluating trajectory-level deletion in offline RL is difficult because lower membership scores can reflect trajectory removal, residual memorization, or policy collapse
PolyGIN architecture with L polynomial transformation blocks has derivative along attribution path with degree at most 2^L-1, enabling exact Gauss-Legendre quadrature evaluation
APEX framework enables exactly computable attribution integrals for graph neural networks through polynomial GNN architecture, eliminating the trade-off between quadrature error and computational cost in path-integral attribution methods
Flash EQ-Linear can reduce computational complexity from O(NDC) to O(NDC/T) for equivariant linear layers
Pipeline data may be stale or degraded.
Last synthesis: 2026-09-20. 8,966 pending.