gankim/tree-of-clarifications

🌲 Code for our EMNLP 2023 paper - πŸŽ„ "Tree of Clarifications: Answering Ambiguous Questions with Retrieval-Augmented Large Language Models"

28
/ 100
Experimental

Implements a tree-based search strategy where LLMs explore multiple interpretations of ambiguous questions and prune low-confidence branches before generating answers. Integrates retrieval-augmented generation via ColBERT dense retrieval and optional Bing Search for external context, with OpenAI's API for language generation. Evaluated on the ASQA benchmark for long-form answer generation with state-of-the-art performance.

No commits in the last 6 months.

No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 8 / 25
Community 12 / 25

How are scores calculated?

Stars

54

Forks

6

Language

Python

License

Last pushed

Dec 04, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/rag/gankim/tree-of-clarifications"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.