Many-shot In-Context Learning
Also called many-shot ICL, long-context ICL.
Many-shot in-context learning places hundreds or thousands of demonstrations in a long-context prompt, often yielding large gains over few-shot prompting.
Description
Agarwal et al. found consistent improvements as the number of shots grew into the hundreds, and introduced variants that use model-generated rationales or unlabeled problems. The same scaling behavior underlies the many-shot jailbreaking attack.
Sources
- Agarwal et al. (2024). Many-Shot In-Context Learning.
Cite this entry
Protologue. (2026). Many-shot In-Context Learning. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.0.0, PTL-0022). https://protologue.com/t/many-shot-in-context-learning/
BibTeX
@misc{protologue_many_shot_in_context_learning,
title = {Many-shot In-Context Learning},
author = {{Protologue}},
year = {2026},
howpublished = {Protologue: A Taxonomy of Prompting and LLM Techniques, v1.0.0},
note = {Entry PTL-0022},
url = {https://protologue.com/t/many-shot-in-context-learning/}
}