Prompting on English AI Terms Dictionary

Zero-Shot Prompting

Sat, 18 Jul 2026 10:20:18 +0000

Definition

Zero-shot prompting involves asking a pre-trained language model to complete a task directly via a textual prompt, without providing any few-shot examples or performing additional fine-tuning. The model leverages its extensive pre-training knowledge to infer the task requirements from the instruction alone. This approach highlights the emergent capabilities of large models, allowing for flexible task adaptation across domains like summarization, classification, and generation with minimal overhead.

Summary

A technique where large language models perform tasks without prior examples or fine-tuning, relying solely on natural language instructions.

Tree of Thoughts

Sat, 18 Jul 2026 10:18:52 +0000

Definition

Tree of Thoughts (ToT) extends traditional chain-of-thought prompting by allowing the model to explore multiple distinct reasoning paths at each step, forming a tree structure. The model evaluates these ’thoughts’ to decide which branches to pursue further, enabling it to look ahead, backtrack from dead ends, and make global planning decisions. This approach significantly improves performance on tasks requiring strategic planning, creative generation, or complex problem-solving.

Summary

Tree of Thoughts is a reasoning framework that explores multiple possible reasoning paths simultaneously, evaluating them to select the most promising next step.

Reasoning model

Sat, 18 Jul 2026 10:13:36 +0000

Definition

Unlike standard generative models focused on fluency, reasoning models prioritize accuracy in multi-step tasks such as mathematics, coding, and logical puzzles. They often employ techniques like Chain-of-Thought prompting or reinforcement learning from logical feedback. These models excel at breaking down ambiguous problems into solvable components, reducing hallucination rates in critical applications requiring strict logical consistency.

Summary

An AI model specifically optimized to perform complex logical deduction, step-by-step problem solving, and chain-of-thought processing.

ReAct

Sat, 18 Jul 2026 09:42:48 +0000

Definition

The ReAct framework enables LLMs to generate both reasoning traces and task-specific actions in an interleaved manner. By simulating human-like thought processes, it allows models to interact with external environments, such as search engines or calculators, to verify facts and solve problems step-by-step. This synergy reduces hallucinations and enhances the accuracy of responses in question-answering and planning scenarios.

Summary

ReAct is a prompting paradigm that combines reasoning and acting to improve the performance of large language models on complex tasks.

Few-Shot Prompting

Sat, 18 Jul 2026 09:40:59 +0000

Definition

This method leverages the in-context learning capabilities of large language models by providing a few illustrative examples directly in the prompt. Unlike fine-tuning, which requires updating model weights, few-shot prompting allows users to steer the model’s output format, tone, or logic dynamically. It is highly effective for tasks like classification, translation, or code generation where specific patterns need to be demonstrated to the model before generating the final response.

In-Context Learning

Sat, 18 Jul 2026 07:39:00 +0000

Definition

In-context learning (ICL) allows large language models to adapt to new tasks without updating their weights. By providing input-output pairs within the prompt context, the model infers the pattern and applies it to new queries. This zero-shot or few-shot capability enables rapid prototyping and flexibility, serving as a powerful alternative to traditional fine-tuning for tasks requiring quick adaptation to novel domains.

Summary

A technique where models learn to perform tasks by observing examples provided in the prompt.

Chain-of-Thought

Sat, 18 Jul 2026 07:38:30 +0000

Definition

Chain-of-Thought (CoT) prompting is a strategy where large language models are guided to produce step-by-step reasoning explanations before arriving at a final answer. By breaking down complex problems into intermediate logical steps, CoT enhances the model’s ability to handle arithmetic, commonsense, and symbolic reasoning tasks. This technique leverages the model’s latent reasoning capabilities, significantly improving accuracy and interpretability in solving multi-step problems.

Summary

A prompting technique that encourages LLMs to generate intermediate reasoning steps before answering.