Today for AI

EXPLORE / 06

Terms index

Specialized AI and software-development terms with concise explanations

Entity records

4517 records · 80/113

  1. 3161

    I/O Bottleneck

    Performance limitation caused by slower input/output operations than computation

    Website pending
  2. 3162

    Streaming Loading

    Reading data chunks from remote storage on demand to avoid full downloads

    Website pending
  3. 3163

    Inference

    The process of using a trained model to make predictions or generate outputs on new data

    Website pending
  4. 3164

    Policy Training

    The process in reinforcement learning where an agent learns decision rules to maximize rewards

    Website pending
  5. 3165

    Digital Twin

    A virtual replica of a physical entity used for simulation, analysis, and optimization

    Website pending
  6. 3166

    Cambricon

    Leading Chinese intelligent chip design company providing cloud and terminal AI acceleration solutions

    Website pending
  7. 3167

    Ascend

    Huawei's full-scenario AI processor series used for cloud training and edge inference

    Website pending
  8. 3168

    Agent Alignment

    Training process ensuring AI agent behaviors align with human intent and task constraints

    Website pending
  9. 3169

    Model Garden

    A centralized catalog and management interface for pre-trained models provided by cloud platforms.

    Website pending
  10. 3170

    Mamba Architecture

    Efficient sequence modeling architecture based on State Space Models

    Website pending
  11. 3171

    Nix

    A declarative package manager ensuring consistent and reproducible build environments

    Website pending
  12. 3172

    ROCm

    AMD's open-source GPU computing platform, comparable to NVIDIA CUDA

    Website pending
  13. 3173

    API Abstraction Layer

    Middleware pattern hiding backend differences behind a unified interface

    Website pending
  14. 3174

    Core ML

    Apple's framework for running machine learning models on-device

    Website pending
  15. 3175

    Long-Form Transcription

    Transcribing audio files with extended duration and complex context

    Website pending
  16. 3176

    Pre-training from Scratch

    The process of training a model directly on large-scale data without relying on existing weights.

    Website pending
  17. 3177

    Cloud GPU Orchestration

    Automated allocation and management of remote graphics processing units for compute-intensive tasks

    Website pending
  18. 3178

    Standardized Interface

    Specifications unifying interactions between modules to enhance compatibility and maintainability.

    Website pending
  19. 3179

    Model Architecture

    The structural design of a neural network determining parameter connections and computational logic.

    Website pending
  20. 3180

    Code Sandbox

    An isolated runtime environment used to safely execute generated code to prevent malicious operations.

    Website pending
  21. 3181

    Code Interpreter

    A tool component allowing LLMs to generate and execute code to solve complex problems.

    Website pending
  22. 3182

    Model Hub

    Online platform for storing, versioning, and distributing pre-trained models

    Website pending
  23. 3183

    Multi-process Architecture

    Design running tasks in separate processes to enhance stability and isolation

    Website pending
  24. 3184

    Inference Server

    Software component that receives requests and runs models to generate outputs

    Website pending
  25. 3185

    Guardrail Model

    Auxiliary models designed to detect potential risks and safety violations in inputs or outputs

    Website pending
  26. 3186

    Predefined Policies

    Pre-set rules or constraints used to guide model behavioral boundaries.

    Website pending
  27. 3187

    Vocabulary

    Set of all possible input tokens defined by the model

    Website pending
  28. 3188

    Chat Completion

    A large model text generation interface paradigm based on turn-based conversations.

    Website pending
  29. 3189

    Reasoning LLM

    Large language models capable of logical deduction and multi-step planning

    Website pending
  30. 3190

    Hybrid Architecture

    Model design combining multiple neural components or attention mechanisms to balance efficiency and performance

    Website pending
  31. 3191

    Inference Standard

    Unified technical standards regulating model input/output formats and interaction protocols.

    Website pending
  32. 3192

    Video Diffusion Model

    Deep learning architecture that generates video frames by progressively denoising from random noise.

    Website pending
  33. 3193

    Differential Attention

    A technique that suppresses noise and enhances signals by computing the difference between two sets of attention scores.

    Website pending
  34. 3194

    GQA

    An attention mechanism where multiple query heads share a single set of key-value heads, balancing between MHA and MQA.

    Website pending
  35. 3195

    Asset Operations

    The process of monitoring, maintaining, and managing physical or digital assets

    Website pending
  36. 3196

    Agent Benchmark

    Standardized test sets used to evaluate AI agent performance on specific tasks

    Website pending
  37. 3197

    Open-source LLM

    Large language models with public weights allowing free community modification and use

    Website pending
  38. 3198

    On-policy Collection

    Action sequences generated by the current policy used to update model parameters

    Website pending
  39. 3199

    Dialect Processing

    NLU and generation techniques for specific regional language variants

    Website pending
  40. 3200

    On-policy

    Training method using data generated by the current policy being optimized

    Website pending