Skip to content

Embedding

Definition
AI-generated

An embedding is a learned, fixed- or variable-length vector (or low-rank factor) representing a discrete token, span, image patch, cell, or other object so that semantically similar items map near each other in the space; modern transformers compute contextual embeddings that depend on surrounding context.

Topics
LLM and Agents Machine learning concepts
Related
Genome-Wide Association Study (GWAS) Large Language Model (LLM) Multimodal Model

Why it matters in GWAS

Sequence and single-cell models use embeddings as inputs to heads for variant scoring or cell typing; GWAS interpretability usually requires independent association evidence because embedding geometry can absorb confounding or batch signal.

Example usage

"The variant window was tokenized into k-mers and passed through a pretrained DNA encoder to obtain a pooled embedding for the MLP risk head."

Genome-Wide Association Study (GWAS) Large Language Model (LLM) Multimodal Model Neural Network Positional Encoding Token Tokenization Transformer
Related-term network

Click any node to open that term page.

Node order. The current term is first. Next come every page linked from Related terms, in the same top-to-bottom order as that list. Then new terms appear in two-hop layers: for each of those direct neighbors (in that same order), their related terms are appended if not already listed. A third pass does the same for two-hop nodes. The walk stops when a fixed maximum number of nodes is reached.

References

  • Mikolov T, et al. (2013). Distributed representations of words and phrases and their compositionality. NeurIPS.
  • Devlin J, et al. (2019). BERT: pre-training of deep bidirectional transformers for language understanding. NAACL.
← Electronic Health Records (EHRs) Encyclopedia of DNA Elements (ENCODE) →

Last updated (UTC · Git history)