What is In-context learning(ICL) in LLMs

Let's talk about an interesting way LLMs can be used after their training – it's called In-Context Learning (ICL). Basically, the model can learn to tackle a new task on the fly, without needing any updates to its training data. When you give it a prompt with some examples related to the task, it can handle anything from translation and speech synthesis to sentiment analysis and creating content.

The magic of LLMs lies in their knack for understanding and reproducing patterns in written language. This means they can generate text that makes sense in a given context. It's like they have this special ability to read between the lines and create text that fits the situation.

What is in-context learning (ICL)?

Traditional machine learning models were primarily designed to tackle specific tasks based on their training data. Their capabilities were bound by the input-output pairs they were trained on, and any deviation from this would lead to suboptimal results. However, with the emergence of LLMs, a paradigm shift occurred in how we solved natural language tasks.

In-context learning (ICL) is a technique where task demonstrations are integrated into the prompt in a natural language format. This approach allows pre-trained LLMs to address new tasks without fine-tuning the model.

Unlike supervised learning, which necessitates a training phase involving backpropagation to modify model parameters, ICL operates without updating these parameters and executes predictions using pre-trained language models. The model determines the underlying patterns within the provided latent space and generates accurate predictions accordingly.

In-context learning (ICL) is known as few-shot learning or few-shot prompting. Contrary to conventional models, the knowledge accumulated via this method is transient; post-inference, the LLM does not persistently store this information, ensuring the stability of model parameters.

ICL's efficacy is attributed to its capacity to exploit the extensive pre-training data and the expansive model scale inherent to LLMs. This allows LLMs to comprehend and execute novel tasks without a comprehensive training process of preceding machine learning architectures.