EXPLORE / 06
Terms index
Specialized AI and software-development terms with concise explanations
RECORDS
Entity records
3387 records · 61/85
- 2401
Q-Ensemble
Training multiple independent Q-value networks to estimate the distribution and uncertainty of the value function
Website pending - 2402
UCB
An exploration strategy based on optimism bias that prioritizes actions with the highest potential payoff
Website pending - 2403
Robotic Simulation
The technique of simulating robot dynamics in virtual environments for testing and training
Website pending - 2404
MuJoCo
A high-performance physics engine for computational physics, machine learning, and biomechanics
Website pending - 2405
Adversarial Examples
Inputs with small perturbations that cause models to produce incorrect outputs.
Website pending - 2406
Curriculum Learning
A training strategy that arranges samples from easy to hard, mimicking human learning.
Website pending - 2407
Hindsight
Re-defining goals based on actual outcomes to generate valid samples
Website pending - 2408
Experience Replay
Storing past interactions and sampling them randomly to break correlations
Website pending - 2409
Proximal Policy Optimization
A reinforcement learning algorithm that stabilizes training by limiting policy update steps
Website pending - 2410
Exploration-Exploitation
The trade-off between trying new actions to gather information and choosing known best actions.
Website pending - 2411
Reward Function
Numerical scoring mechanism defining agent behavior quality in reinforcement learning
Website pending - 2412
Supervised Learning
A machine learning paradigm using labeled data to train models for prediction.
Website pending - 2413
Sample Efficiency
Higher efficiency means fewer data points or interactions are needed to achieve specific performance.
Website pending - 2414
TRPO
An algorithm ensuring monotonic improvement by limiting the step size of policy updates.
Website pending - 2415
A3C
A reinforcement learning algorithm that updates policies using multiple parallel worker threads.
Website pending - 2416
Supervised Deep Learning
Model training method relying on labeled static datasets.
Website pending - 2417
Theory of Mind
The ability to infer others' beliefs, intentions, and mental states, often used to describe AI modeling of interaction partners.
Website pending - 2418
Opponent Modeling
Process of predicting future behaviors or policy changes of other agents
Website pending - 2419
Opponent-Learning Awareness
Method where an agent explicitly models the future learning behavior of opponents during policy updates
Website pending - 2420
Deep Linear Network
Neural network composed of multiple linear layers without activation functions
Website pending - 2421
Autoregressive Model
A generative architecture that predicts subsequent sequence elements based on prior outputs.
Website pending - 2422
Visual Reinforcement Learning
Reinforcement learning methods that learn decision-making policies directly from image pixel inputs.
Website pending - 2423
Dynamics Randomization
A method enhancing policy robustness by randomly varying physical parameters during training
Website pending - 2424
Asymmetric Architecture
Network design where different components use distinct structures or information levels
Website pending - 2425
Actor-Critic
A reinforcement learning framework combining policy gradients with value estimation
Website pending - 2426
High-level Action
Abstract behavioral units composed of low-level policies, used to simplify decision spaces
Website pending - 2427
Closed-loop Control
Control system that uses feedback to adjust outputs in real time to maintain stability or achieve goals
Website pending - 2428
Sim-to-Real
Technique of transferring policies trained in simulated environments to physical reality
Website pending - 2429
Long-Horizon Dependency
A property of sequential problems where current decisions significantly impact distant future states.
Website pending - 2430
Hierarchical Reinforcement Learning
An RL method that decomposes tasks into subgoals or high-level actions to simplify decision-making.
Website pending - 2431
L0 Regularization
A mathematical method that enforces model sparsity by minimizing the number of non-zero elements
Website pending - 2432
Sparse Neural Network
A network structure where most connection weights are zero, reducing computation and storage needs
Website pending - 2433
Non-exclusive Classification
A classification mechanism allowing a sample to belong to multiple categories simultaneously
Website pending - 2434
Branching Factor
The average number of child nodes per node in a search tree, affecting computational complexity.
Website pending - 2435
Ontology
A formal specification of concepts and their relationships, often used in knowledge graph construction.
Website pending - 2436
Entity Disambiguation
The process of determining which unique object in a knowledge base is referred to by an entity mention in text.
Website pending - 2437
CUDA kernel
Core parallel computation functions executed on NVIDIA GPUs
Website pending - 2438
Block-sparse matrix
A sparse data structure where non-zero elements are distributed in fixed-size blocks
Website pending - 2439
Knowledge Distillation
Technique transferring knowledge from a large model to a smaller or different one
Website pending - 2440
Hindsight Experience Replay
A technique that relabels failed trajectories as successes for other goals.
Website pending