Learn Before
Improving Chain-of-Thought Reasoning
The accuracy of chain-of-thought reasoning in language models can be enhanced through various strategic modifications. These include sampling multiple distinct reasoning paths to find consensus (self-consistency), diversifying the provided few-shot demonstrations, and reducing complex problems into smaller, manageable sub-problems.
0
1
Tags
D2L
Dive into Deep Learning @ D2L
Related
Example of a Model Initiating Chain-of-Thought Reasoning
A user is trying to solve a word problem using a large language model. They use two different prompts and receive two different responses.
Prompt 1: "A farmer has 5 pens, and each pen holds 8 chickens. The farmer buys 10 more chickens. How many chickens does the farmer have in total?" Response 1: "The farmer has 50 chickens in total."
Prompt 2: "A farmer has 5 pens, and each pen holds 8 chickens. The farmer buys 10 more chickens. How many chickens does the farmer have in total? Let's think step by step." Response 2: "First, we find the initial number of chickens. The farmer has 5 pens with 8 chickens each, so 5 * 8 = 40 chickens. Then, the farmer buys 10 more chickens. So, we add those to the initial amount: 40 + 10 = 50 chickens. The farmer has 50 chickens in total."
Based on the information provided, what is the most likely reason for the difference in the structure and detail of the two responses?
Modifying a Prompt for Step-by-Step Reasoning
Improving AI Reasoning for a Multi-Step Problem
Improving Chain-of-Thought Reasoning
Zero-Shot Chain-of-Thought Reasoning
Multimodal Chain-of-Thought Reasoning