# In-Context Learning Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/in-context-learning
> Markdown URL: https://aitinkerers.org/technologies/in-context-learning.md
> Technology record last updated: 2026-02-26T13:03:07Z
> Generated: 2026-09-21T23:39:51Z

LLMs adapt to a new task by leveraging examples within the input prompt (few-shot prompting), eliminating costly parameter updates.

In-Context Learning (ICL) is an emergent Large Language Model (LLM) capability: it allows models like GPT-3 to perform novel tasks based on demonstrations provided directly in the prompt (e.g., zero-, one-, or few-shot examples). The core mechanism involves conditioning the pre-trained model on this temporary context, which guides the output without requiring backpropagation or model weight updates. This approach delivers rapid, flexible task adaptation, significantly reducing the time and computational resources associated with traditional fine-tuning.

- Official technology site: https://www.ibm.com/topics/in-context-learning
- Public AI Tinkerers demos and talks: 2
- Result page: 1 of 1

## Recent Public Talks and Demos

### [GESTURE (Graphical Encoding System for Translating Unique Recognisable Expressions)](https://lausanne.aitinkerers.org/talks/rsvp_9R4zaMzbav4)

The GESTURE project aims to solve a sign language translation task by translating a sequence of ASL signs into natural language. By connecting a live camera feed to a Pose Extractor (Google MediaPipe), signs can be detected in real time and translated/classified into their associated word(s) in natural language. Words are then translated into full sentences using an In-Context learning prompt sent to ChatGPT.

- Event context: AI Tinkerers Lausanne June 2025 Meetup — 2025-06-16 — Lausanne
- Public talk page: https://lausanne.aitinkerers.org/talks/rsvp_9R4zaMzbav4

### [In-Context Learning for Extreme Multi-Label Classification](https://chicago.aitinkerers.org/talks/rsvp_y-7-uutZyio)

How to leverage large language models to identify > 10,000 different labels with < 10 labelled examples for each class? And how to do it by systematically searching the prompt space without tinkering with prompts for each class? There is this Stanford paper (From Chris Potts' group) and an accompanying Github repository that offers an answer to the questions above. I have been tinkering with it. The idea was to share what I am learning with fellow tinkerers in a quick talk.

- Event context: AI Tinkerers Chicago July Meetup — 2024-07-23 — Chicago
- Public talk page: https://chicago.aitinkerers.org/talks/rsvp_y-7-uutZyio

## Related Technologies

- [Apache Spark](https://aitinkerers.org/technologies/apache-spark) ([Markdown](https://aitinkerers.org/technologies/apache-spark.md)) — 1 public demo
- [BERT](https://aitinkerers.org/technologies/bert) ([Markdown](https://aitinkerers.org/technologies/bert.md)) — 179 public demos
- [BLOOM](https://aitinkerers.org/technologies/bloom) ([Markdown](https://aitinkerers.org/technologies/bloom.md)) — 115 public demos
- [ChatGPT 4o Mini](https://aitinkerers.org/technologies/chatgpt-4o-mini) ([Markdown](https://aitinkerers.org/technologies/chatgpt-4o-mini.md)) — 1 public demo
- [ECC](https://aitinkerers.org/technologies/ecc) ([Markdown](https://aitinkerers.org/technologies/ecc.md)) — 1 public demo
- [fastXML](https://aitinkerers.org/technologies/fastxml) ([Markdown](https://aitinkerers.org/technologies/fastxml.md)) — 1 public demo
- [GitHub](https://aitinkerers.org/technologies/github) ([Markdown](https://aitinkerers.org/technologies/github.md)) — 74 public demos
- [Google MediaPipe](https://aitinkerers.org/technologies/google-mediapipe) ([Markdown](https://aitinkerers.org/technologies/google-mediapipe.md)) — 1 public demo
- [GPT-3](https://aitinkerers.org/technologies/gpt-3) ([Markdown](https://aitinkerers.org/technologies/gpt-3.md)) — 191 public demos
- [GPT-4](https://aitinkerers.org/technologies/gpt-4) ([Markdown](https://aitinkerers.org/technologies/gpt-4.md)) — 529 public demos
- [Llama-2](https://aitinkerers.org/technologies/llama-2) ([Markdown](https://aitinkerers.org/technologies/llama-2.md)) — 227 public demos
- [Mulan](https://aitinkerers.org/technologies/mulan) ([Markdown](https://aitinkerers.org/technologies/mulan.md)) — 1 public demo
- [OpenAI API](https://aitinkerers.org/technologies/openai-api) ([Markdown](https://aitinkerers.org/technologies/openai-api.md)) — 520 public demos
- [PaLM 2](https://aitinkerers.org/technologies/palm-2) ([Markdown](https://aitinkerers.org/technologies/palm-2.md)) — 116 public demos
- [prompt space](https://aitinkerers.org/technologies/prompt-space) ([Markdown](https://aitinkerers.org/technologies/prompt-space.md)) — 1 public demo
- [RAkEL](https://aitinkerers.org/technologies/rakel) ([Markdown](https://aitinkerers.org/technologies/rakel.md)) — 1 public demo
- [RoBERTa](https://aitinkerers.org/technologies/roberta) ([Markdown](https://aitinkerers.org/technologies/roberta.md)) — 118 public demos
- [scikit-learn](https://aitinkerers.org/technologies/scikit-learn) ([Markdown](https://aitinkerers.org/technologies/scikit-learn.md)) — 84 public demos