Few Shot Learning - Revision history

BPeat at 01:15, 19 September 2026

2026-09-19T01:15:14Z

BPeat at 21:04, 16 September 2026

2026-09-16T21:04:41Z

BPeat at 20:37, 28 April 2024

2024-04-28T20:37:35Z

BPeat: /* Zero-Shot Prompting */

2023-09-27T11:41:58Z

‎Zero-Shot Prompting

BPeat: /* Zero-Shot Prompting */

2023-09-27T11:40:05Z

‎Zero-Shot Prompting

BPeat: /* Zero-shot Chain of Thought (CoT) prompting */

2023-09-27T11:37:51Z

‎Zero-shot Chain of Thought (CoT) prompting

BPeat: /* Zero-Shot Learning */

2023-09-27T11:36:42Z

‎Zero-Shot Learning

BPeat: /* Zero-Shot Learning */

2023-09-27T11:34:44Z

‎Zero-Shot Learning

BPeat: /* Zero-Shot Learning */

2023-09-27T11:34:04Z

‎Zero-Shot Learning

BPeat at 15:49, 13 September 2023

2023-09-13T15:49:43Z

@@ Line 16: / Line 16: @@
 * [[Prompt Engineering (PE)]] ...[[Prompt Engineering (PE)#PromptBase|PromptBase]] ... [[Prompt Injection Attack]]
 * [[Mixture-of-Experts (MoE)]] ... [[Chain of Thought (CoT)]] ... [[Chain of Thought (CoT)#Tree of Thoughts (ToT)|Tree of Thoughts (ToT)]] ... [[Artificial_General_Intelligence_(AGI)_to_Singularity#Theory%20of%20Mind%20(ToM)|Theory of Mind (ToM)]]
-* [[Embedding]] ... [[Fine-tuning]] ... [[Retrieval-Augmented Generation (RAG)|RAG]] ... [[Agents#AI-Powered Search|Search]] ... [[Clustering]] ... [[Recommendation]] ... [[Anomaly Detection]] ... [[Classification]] ... [[Dimensional Reduction]].  [[...find outliers]]
+* [[Embedding]] ... [[Fine-tuning]] ... [[Retrieval-Augmented Generation (RAG)|RAG]] ... [[Clustering]] ... [[Recommendation]] ... [[Anomaly Detection]] ... [[Classification]] ... [[Dimensional Reduction]].  [[...find outliers]]
 * [https://medium.com/quick-code/understanding-few-shot-learning-in-machine-learning-bede251a0f67 Understanding few-shot learning in machine learning | Michael J. Garbade]
 * [https://arxiv.org/pdf/1803.02999.pdf On First-Order Meta-Learning Algorithms | A. Nichol, J. Achiam,  and J. Schulman -] [[OpenAI]]

@@ Line 15: / Line 15: @@
 * [[Causation vs. Correlation]] ... [[Autocorrelation]] ...[[Convolution vs. Cross-Correlation (Autocorrelation)]]
 * [[Prompt Engineering (PE)]] ...[[Prompt Engineering (PE)#PromptBase|PromptBase]] ... [[Prompt Injection Attack]]
 * [[Embedding]] ... [[Fine-tuning]] ... [[Retrieval-Augmented Generation (RAG)|RAG]] ... [[Agents#AI-Powered Search|Search]] ... [[Clustering]] ... [[Recommendation]] ... [[Anomaly Detection]] ... [[Classification]] ... [[Dimensional Reduction]].  [[...find outliers]]
 * [https://medium.com/quick-code/understanding-few-shot-learning-in-machine-learning-bede251a0f67 Understanding few-shot learning in machine learning | Michael J. Garbade]

@@ Line 12: / Line 12: @@
 * [[Learning Techniques]]
-* [[In-Context Learning (ICL)]] ... [[Context]] ... [[Causation vs. Correlation]] ... [[Autocorrelation]] ... [[Out-of-Distribution (OOD) Generalization]] ... [[Transfer Learning]]
+* [[Perspective]] ... [[Context]] ... [[In-Context Learning (ICL)]] ... [[Transfer Learning]] ... [[Out-of-Distribution (OOD) Generalization]]
 * [[Prompt Engineering (PE)]] ...[[Prompt Engineering (PE)#PromptBase|PromptBase]] ... [[Prompt Injection Attack]]
 * [[Embedding]] ... [[Fine-tuning]] ... [[Retrieval-Augmented Generation (RAG)|RAG]] ... [[Agents#AI-Powered Search|Search]] ... [[Clustering]] ... [[Recommendation]] ... [[Anomaly Detection]] ... [[Classification]] ... [[Dimensional Reduction]].  [[...find outliers]]

@@ Line 27: / Line 27: @@
 ==== <span id="Zero-Shot Prompting"></span>Zero-Shot Prompting ====
-Zero-Shot [[Chain of Thought (CoT)]] prompting is a technique for enabling [[Large Language Model (LLM)]]s to perform multi-step reasoning without any training on task-specific examples through [[Fine-tuning#Implementing Chain of Thought (CoT)|Fine-tuning. This is in contrast to traditional [[Chain of Thought (CoT)|CoT]] prompting, which requires a few hand-crafted step-by-step reasoning demonstrations for each task. Zero-shot [[Chain of Thought (CoT)|CoT]] prompting is possible because [[Large Language Model (LLM)|LLM]]s are trained on massive datasets of text and code, which gives them a broad understanding of the world and the ability to perform many different kinds of tasks. When prompted with a zero-shot [[Chain of Thought (CoT)|CoT]] prompt, an [[Large Language Model (LLM)|LLM]] can use its knowledge and reasoning abilities to solve the task, even if it has never seen a similar task before.
+Zero-Shot [[Chain of Thought (CoT)]] prompting is a technique for enabling [[Large Language Model (LLM)]]s to perform multi-step reasoning without any training on task-specific examples through [[Fine-tuning#Implementing Chain of Thought (CoT)|Fine-tuning]]. This is in contrast to traditional [[Chain of Thought (CoT)|CoT]] prompting, which requires a few hand-crafted step-by-step reasoning demonstrations for each task. Zero-shot [[Chain of Thought (CoT)|CoT]] prompting is possible because [[Large Language Model (LLM)|LLM]]s are trained on massive datasets of text and code, which gives them a broad understanding of the world and the ability to perform many different kinds of tasks. When prompted with a zero-shot [[Chain of Thought (CoT)|CoT]] prompt, an [[Large Language Model (LLM)|LLM]] can use its knowledge and reasoning abilities to solve the task, even if it has never seen a similar task before.
 Here are some of the benefits of using zero-shot [[Chain of Thought (CoT)|CoT]] prompting:

@@ Line 26: / Line 26: @@
 A model is trained without using any labeled data. Instead of giving training examples, zero-shot learning gives a high-level description of new categories so that the machine can relate it to existing categories that the machine has learned about. For example, [[Hugging Face]] [[transformer]]s use zero-shot classification for more than 60 per cent of its [[transformer]]s.
-==== Zero-Shot Prompting ====
+==== <span id="Zero-Shot Prompting"></span>Zero-Shot Prompting ====
-Zero-shot [[Chain of Thought (CoT)]] prompting is a technique for enabling [[Large Language Model (LLM)]]s to perform multi-step reasoning without any training on task-specific examples through [[Fine-tuning#Implementing Chain of Thought (CoT)|Fine-tuning. This is in contrast to traditional [[Chain of Thought (CoT)|CoT]] prompting, which requires a few hand-crafted step-by-step reasoning demonstrations for each task. Zero-shot [[Chain of Thought (CoT)|CoT]] prompting is possible because [[Large Language Model (LLM)|LLM]]s are trained on massive datasets of text and code, which gives them a broad understanding of the world and the ability to perform many different kinds of tasks. When prompted with a zero-shot [[Chain of Thought (CoT)|CoT]] prompt, an [[Large Language Model (LLM)|LLM]] can use its knowledge and reasoning abilities to solve the task, even if it has never seen a similar task before.
+Zero-Shot [[Chain of Thought (CoT)]] prompting is a technique for enabling [[Large Language Model (LLM)]]s to perform multi-step reasoning without any training on task-specific examples through [[Fine-tuning#Implementing Chain of Thought (CoT)|Fine-tuning. This is in contrast to traditional [[Chain of Thought (CoT)|CoT]] prompting, which requires a few hand-crafted step-by-step reasoning demonstrations for each task. Zero-shot [[Chain of Thought (CoT)|CoT]] prompting is possible because [[Large Language Model (LLM)|LLM]]s are trained on massive datasets of text and code, which gives them a broad understanding of the world and the ability to perform many different kinds of tasks. When prompted with a zero-shot [[Chain of Thought (CoT)|CoT]] prompt, an [[Large Language Model (LLM)|LLM]] can use its knowledge and reasoning abilities to solve the task, even if it has never seen a similar task before.
 Here are some of the benefits of using zero-shot [[Chain of Thought (CoT)|CoT]] prompting:

@@ Line 24: / Line 24: @@
 === <span id="Zero-Shot Learning"></span>Zero-Shot Learning ===
-A model is trained without using any labeled data. Instead of giving training examples, zero-shot learning gives a high-level description of new categories so that the machine can relate it to existing categories that the machine has learned about. For example, [[Hugging Face]] [[transformer]s use zero-shot classification for more than 60 per cent of its transformers.
+A model is trained without using any labeled data. Instead of giving training examples, zero-shot learning gives a high-level description of new categories so that the machine can relate it to existing categories that the machine has learned about. For example, [[Hugging Face]] [[transformer]]s use zero-shot classification for more than 60 per cent of its [[transformer]]s.
 ==== Zero-shot [[Chain of Thought (CoT)]] prompting ====

← Older revision		Revision as of 11:34, 27 September 2023
Line 26:		Line 26:
	A model is trained without using any labeled data. Instead of giving training examples, zero-shot learning gives a high-level description of new categories so that the machine can relate it to existing categories that the machine has learned about. For example, [[Hugging Face]] [[transformer]s use zero-shot classification for more than 60 per cent of its transformers.		A model is trained without using any labeled data. Instead of giving training examples, zero-shot learning gives a high-level description of new categories so that the machine can relate it to existing categories that the machine has learned about. For example, [[Hugging Face]] [[transformer]s use zero-shot classification for more than 60 per cent of its transformers.

		+	==== Zero-shot [[Chain of Thought (CoT)]] prompting ====
	Zero-shot [[Chain of Thought (CoT)]] prompting is a technique for enabling [[Large Language Model (LLM)]]s to perform multi-step reasoning without any training on task-specific examples through [[Fine-tuning#Implementing Chain of Thought (CoT)\|Fine-tuning. This is in contrast to traditional [[Chain of Thought (CoT)\|CoT]] prompting, which requires a few hand-crafted step-by-step reasoning demonstrations for each task. Zero-shot [[Chain of Thought (CoT)\|CoT]] prompting is possible because [[Large Language Model (LLM)\|LLM]]s are trained on massive datasets of text and code, which gives them a broad understanding of the world and the ability to perform many different kinds of tasks. When prompted with a zero-shot [[Chain of Thought (CoT)\|CoT]] prompt, an [[Large Language Model (LLM)\|LLM]] can use its knowledge and reasoning abilities to solve the task, even if it has never seen a similar task before.		Zero-shot [[Chain of Thought (CoT)]] prompting is a technique for enabling [[Large Language Model (LLM)]]s to perform multi-step reasoning without any training on task-specific examples through [[Fine-tuning#Implementing Chain of Thought (CoT)\|Fine-tuning. This is in contrast to traditional [[Chain of Thought (CoT)\|CoT]] prompting, which requires a few hand-crafted step-by-step reasoning demonstrations for each task. Zero-shot [[Chain of Thought (CoT)\|CoT]] prompting is possible because [[Large Language Model (LLM)\|LLM]]s are trained on massive datasets of text and code, which gives them a broad understanding of the world and the ability to perform many different kinds of tasks. When prompted with a zero-shot [[Chain of Thought (CoT)\|CoT]] prompt, an [[Large Language Model (LLM)\|LLM]] can use its knowledge and reasoning abilities to solve the task, even if it has never seen a similar task before.