Atlas: Few-shot Learning with Retrieval Augmented Language Models
by Meta AI / University College London · free · Last verified 2026-03-17
Presents Atlas, a retrieval-augmented language model pre-trained to support few-shot learning by jointly training the retriever and language model. Atlas achieves strong few-shot performance across knowledge-intensive tasks, outperforming models with orders of magnitude more parameters on MMLU and other benchmarks.
https://arxiv.org/abs/2208.03299 ↗B
B—Above Average
Adoption: B+Quality: AFreshness: BCitations: AEngagement: F
Specifications
- License
- CC BY-NC 4.0
- Pricing
- free
- Capabilities
- few-shot-learning, retrieval-augmentation, joint-training, knowledge-intensive-nlp
- Integrations
- Use Cases
- few-shot-qa, knowledge-intensive-tasks, research
- API Available
- No
- Tags
- rag, few-shot, retrieval, language-model, joint-training
- Added
- 2026-03-17
- Completeness
- 100%
Index Score
65.2Adoption
70
Quality
86
Freshness
60
Citations
80
Engagement
0
Put AI to work for your business
Deploy this paper alongside autonomous AaaS agents that handle tasks end-to-end — no babysitting required.