EXPLORE / 06
Terms index
Specialized AI and software-development terms with concise explanations
Entity records
4609 records · 82/116
- 3241
Data Construction
The process of preparing, cleaning, and formatting datasets for model training
Website pending - 3242
Depth Pruning
Removing redundant layers from neural networks to compress model structure
Website pending - 3243
Code Execution
Running generated code in a runtime environment to verify functionality
Website pending - 3244
Code Execution Verification
Technical approach checking if code produces expected results by running it
Website pending - 3245
TCO
Total cost of ownership covering hardware, software, and operations lifecycle
Website pending - 3246
Sentence Embedding
Mapping text sentences to fixed-dimension vectors to represent semantic meaning
Website pending - 3247
Backward Compatibility
The ability of a new system version to work with data or interfaces from older versions
Website pending - 3248
Package Dependency
Other packages or libraries required for a software package to function correctly
Website pending - 3249
Open Source Robot Learning
A paradigm using public code and datasets to train robot policies
Website pending - 3250
Model Supply Chain Security
Protecting AI models from malicious tampering or contamination throughout their lifecycle
Website pending - 3251
Typer
A Python framework for building high-performance command-line interfaces based on type hints
Website pending - 3252
httpx
A modern Python HTTP client library supporting HTTP/2 and async requests
Website pending - 3253
I/O Bottleneck
Performance limitation caused by slower input/output operations than computation
Website pending - 3254
Streaming Loading
Reading data chunks from remote storage on demand to avoid full downloads
Website pending - 3255
Inference
The process of using a trained model to make predictions or generate outputs on new data
Website pending - 3256
Policy Training
The process in reinforcement learning where an agent learns decision rules to maximize rewards
Website pending - 3257
Digital Twin
A virtual replica of a physical entity used for simulation, analysis, and optimization
Website pending - 3258
Cambricon
Leading Chinese intelligent chip design company providing cloud and terminal AI acceleration solutions
Website pending - 3259
Ascend
Huawei's full-scenario AI processor series used for cloud training and edge inference
Website pending - 3260
Agent Alignment
Training process ensuring AI agent behaviors align with human intent and task constraints
Website pending - 3261
Model Garden
A centralized catalog and management interface for pre-trained models provided by cloud platforms.
Website pending - 3262
Mamba Architecture
Efficient sequence modeling architecture based on State Space Models
Website pending - 3263
Nix
A declarative package manager ensuring consistent and reproducible build environments
Website pending - 3264
ROCm
AMD's open-source GPU computing platform, comparable to NVIDIA CUDA
Website pending - 3265
API Abstraction Layer
Middleware pattern hiding backend differences behind a unified interface
Website pending - 3266
Core ML
Apple's framework for running machine learning models on-device
Website pending - 3267
Long-Form Transcription
Transcribing audio files with extended duration and complex context
Website pending - 3268
Pre-training from Scratch
The process of training a model directly on large-scale data without relying on existing weights.
Website pending - 3269
Cloud GPU Orchestration
Automated allocation and management of remote graphics processing units for compute-intensive tasks
Website pending - 3270
Standardized Interface
Specifications unifying interactions between modules to enhance compatibility and maintainability.
Website pending - 3271
Model Architecture
The structural design of a neural network determining parameter connections and computational logic.
Website pending - 3272
Code Sandbox
An isolated runtime environment used to safely execute generated code to prevent malicious operations.
Website pending - 3273
Code Interpreter
A tool component allowing LLMs to generate and execute code to solve complex problems.
Website pending - 3274
Model Hub
Online platform for storing, versioning, and distributing pre-trained models
Website pending - 3275
Multi-process Architecture
Design running tasks in separate processes to enhance stability and isolation
Website pending - 3276
Inference Server
Software component that receives requests and runs models to generate outputs
Website pending - 3277
Guardrail Model
Auxiliary models designed to detect potential risks and safety violations in inputs or outputs
Website pending - 3278
Predefined Policies
Pre-set rules or constraints used to guide model behavioral boundaries.
Website pending - 3279
Vocabulary
Set of all possible input tokens defined by the model
Website pending - 3280
Chat Completion
A large model text generation interface paradigm based on turn-based conversations.
Website pending