Papers, with perspective.
What the authors propose. Why the idea matters. What the inspected source cannot establish.
7 opening notes · original abstracts and metadata inspected · English editorial interpretation
Explore connections ↗retrieval · arXiv:1911.00172
Generalization through Memorization: Nearest Neighbor Language Models
A language model consults nearby examples as well as its learned parameters.
retrieval · arXiv:2002.08909
REALM: Retrieval-Augmented Language Model Pre-Training
Retrieval becomes part of language-model pretraining rather than only a downstream add-on.
retrieval · arXiv:2004.04906
Dense Passage Retrieval for Open-Domain Question Answering
A dual encoder retrieves candidate passages using learned dense representations.
retrieval · arXiv:2005.11401
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
A pretrained sequence generator uses a searchable external memory.
retrieval · arXiv:2310.11511
Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection
Learned reflection tokens help control retrieval and critique generation.
retrieval · arXiv:2404.16130
From Local to Global: A Graph RAG Approach to Query-Focused Summarization
Graph communities and summaries support questions about an entire corpus.
systems · arXiv:1706.03762
Attention Is All You Need
The Transformer replaces recurrence and convolution with an attention-based architecture.
Have another paper in mind? The Paper desk accepts arXiv papers on any topic. Importing one there does not add it to this published collection.