EXPLORE / 06
Terms index
Specialized AI and software-development terms with concise explanations
Entity records
4556 records · 57/114
- 2241
Feed-Forward Network (FFN)
Fully connected layer structure in Transformers used for non-linear feature transformation
Website pending - 2242
Parallel Drafting
A strategy generating multiple candidate tokens simultaneously rather than sequentially via autoregression.
Website pending - 2243
SVE256
An implementation of Arm's Scalable Vector Extension instruction set for parallel data processing.
Website pending - 2244
INT8 Quantization
Converting model weights and activations from floating-point to 8-bit integers to accelerate inference.
Website pending - 2245
Day-0 Support
The capability to run a new model seamlessly on specific frameworks or tools on its release day
Website pending - 2246
AllGather
Collective communication primitive gathering data from all nodes and broadcasting to each
Website pending - 2247
mmap
Maps files directly into process virtual address space to avoid redundant copying
Website pending - 2248
Diffusion RL
Technique using reinforcement learning algorithms to optimize diffusion model generation quality
Website pending - 2249
Confidence Head
An additional module in neural networks that outputs reliability scores for predictions.
Website pending - 2250
Non-associativity of Floating Point
The property where changing the order of floating-point additions yields slightly different results due to rounding errors.
Website pending - 2251
Sharded Transfer
Technology that splits large datasets or model weights into smaller chunks for parallel transmission
Website pending - 2252
Online RL
A learning paradigm where agents update policies simultaneously while interacting with the environment
Website pending - 2253
LLM Serving
The technical system for deploying trained large models as online API services.
Website pending - 2254
Prefill/Decode Disaggregation
Separates compute-intensive prefill stages from memory-intensive decode stages.
Website pending - 2255
Preemption
Interrupting current requests to release resources for later reprocessing.
Website pending - 2256
Disaggregated Serving
Architecture separating prefill and decode phases onto different hardware resources to boost throughput
Website pending - 2257
Mixed Precision Quantization
Using different numerical bit-widths across layers or components to balance accuracy and speed.
Website pending - 2258
Bitwise Consistency
Different compute engines producing identical binary outputs for the same inputs.
Website pending - 2259
RL Post-training
The phase of using reinforcement learning to optimize alignment or reasoning after pretraining.
Website pending - 2260
CUDA Operator
High-performance parallel computing kernel functions written based on the NVIDIA CUDA platform.
Website pending - 2261
DSpark
An enhanced block-level speculative decoding algorithm proposed by DeepSeek, aiming to improve token coherence and acceptance length.
Website pending - 2262
Hardware Decoding
Technology using dedicated chip circuits rather than general-purpose CPUs to decompress video streams.
Website pending - 2263
PD Disaggregation
An architecture that deploys prefill and decode phases of LLM inference on separate compute resources to optimize throughput and latency.
Website pending - 2264
Time to First Token (TTFT)
Latency from request initiation to first token generation, measuring responsiveness
Website pending - 2265
Paged Attention
Manages KV cache dynamically using virtual memory principles to eliminate fragmentation
Website pending - 2266
Output Distribution
The set of probability predictions for the next token, reflecting language modeling capability
Website pending - 2267
Adaptive Thinking
A technique where the model dynamically adjusts internal reasoning depth or computational resource allocation based on task complexity.
Website pending - 2268
Runtime Configuration
The ability to dynamically adjust parameters or components during program execution rather than fixing them at compile time.
Website pending - 2269
CUDA Graphs
Technology capturing GPU operation sequences as static graphs to reduce launch overhead
Website pending - 2270
Hot Reloading
Dynamically updating code or configuration without restarting services or interrupting sessions
Website pending - 2271
Agent Trajectory
Sequence of states, actions, and feedback recorded during an agent's task execution
Website pending - 2272
Capability Commoditization
The process where once-scarce advanced AI functions become low-cost generic services as technology spreads
Website pending - 2273
Marginal Value of Intelligence
The additional economic benefit generated by each incremental unit of frontier AI capability
Website pending - 2274
NAND Flash
Non-volatile storage technology widely used in SSDs and mobile devices, characterized by high density.
Website pending - 2275
DRAM
Dynamic Random Access Memory, the primary form of volatile computer main memory requiring periodic refresh.
Website pending - 2276
Subliminal Learning
Behavioral changes induced by implicit signals or indirect data transmission
Website pending - 2277
Self-Modeling
Internal representation and cognition of a model's own capabilities, states, and outputs
Website pending - 2278
Emergent Misalignment
Unexpected harmful behaviors emerging despite good performance on specific tasks
Website pending - 2279
Proprietary Data
Exclusive datasets owned by entities, often used to train models with unique advantages in vertical domains
Website pending - 2280
Quantum Advantage
The capability of a quantum computer to significantly outperform the best classical computer on specific tasks
Website pending