RESEARCH BASE
Search 3,721 documents across 34 fields — every claim tier-rated by evidence
3,633 are the core, quality-scored corpus (34 lettered sections — see How We Work); the remaining 88 are cross-corpus synthesis documents (68 InterDocs, 12 Connections, 8 Theories) also indexed here.
2 results for "self-attention"
ZD_2_15 — Transformer Architecture: Self-Attention and the Foundation of Modern AI
The transformer is a neural network architecture introduced in 2017 that replaced recurrent and convolutional models as the dominant paradigm in artificial intelligence. Its core innovation — the self-attention mechanism
S_1_16 — Large Language Models: Architecture, Capabilities, and Societal Impact
Large Language Models (LLMs) are neural networks with billions to trillions of parameters, trained on massive text corpora to predict the next token in a sequence. Built on the transformer architecture introduced by Vasw
BROWSE BY SECTION — 3,721 documents across 34 fields