Glossary term
Glossary term
Agentic Systems
A sophisticated prompting strategy that encourages an LLM to pursue and refine the most promising intermediate solutions and to abandon the rest. Tree-of-thought prompting uses an algorithm like the following:
Divide a complex problem into different branches (potential strategies), each comprised of multiple steps.
Prompt the LLM to work on each branch independently.
Ask the LLM to evaluate the quality of the solution to each branch after each step.
Continue refining the most promising branch(es); abandon the rest.
If a promising branch eventually fails, backtrack and try other promising steps.
Created for this library
A research-assistant agent uses tree-of-thought prompting to explore several reasoning paths in parallel before settling on a final answer.
A math tutoring tool uses tree-of-thought prompting to consider multiple solution approaches and pick the most promising.
A coding assistant uses tree-of-thought prompting to consider multiple refactor strategies before applying any change.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License