ML Research Wiki / Benchmarks / Question Answering / HotpotQA

HotpotQA

Question Answering Benchmark

Performance Over Time

📊 Showing 22 results | 📏 Metric: JOINT-F1

Top Performing Models

Rank Model Paper JOINT-F1 Date Code
1 Beam Retrieval End-to-End Beam Retrieval for Multi-Hop Question Answering 0.78 2023-08-17 📦 ShayekhBinIslam/openrag 📦 canghongjian/beam_retriever 📦 Alab-NII/2wikimultihop
2 BigBird-etc Big Bird: Transformers for Longer Sequences 0.74 2020-07-28 📦 huggingface/transformers 📦 tensorflow/models 📦 PaddlePaddle/PaddleNLP
3 AISO Adaptive Information Seeking for Open-Domain Question Answering 0.72 2021-09-14 📦 zycdev/aiso
4 Chain-of-Skills Chain-of-Skills: A Configurable Model for Open-domain Question Answering 0.72 2023-05-04 📦 mayer123/udt-qa
5 HopRetriever + Sp-search HopRetriever: Retrieve Hops over Wikipedia to Answer Complex Questions 0.71 2020-12-31 -
6 IRRR+ Answering Open-Domain Questions of Varying Reasoning Steps from Text 0.70 2020-10-23 📦 beerqa/irrr
7 IRRR Answering Open-Domain Questions of Varying Reasoning Steps from Text 0.69 2020-10-23 📦 beerqa/irrr
8 Recursive Dense Retriever Answering Complex Open-Domain Questions with Multi-Hop Dense Retrieval 0.67 2020-09-27 📦 facebookresearch/multihop_dense_retrieval
9 DDRQA Answering Any-hop Open-domain Questions with Iterative Document Reranking 0.64 2020-09-16 -
10 Robustly Fine-tuned Graph-based Recurrent Retriever Learning to Retrieve Reasoning Paths over Wikipedia Graph for Question Answering 0.61 2019-11-24 📦 AkariAsai/learning_to_retrieve_reasoning_paths 📦 AkariAsai/XORQA

All Papers (22)

A Simple Yet Strong Pipeline for HotpotQA

2020
Quark + SemanticRetrievalMRS IR