MultiRC

Multi-Sentence Reading Comprehension

Dataset Information
Modalities
Texts
Languages
English
Introduced
2018
License
Custom (research-only)
Homepage

Overview

MultiRC (Multi-Sentence Reading Comprehension) is a dataset of short paragraphs and multi-sentence questions, i.e., questions that can be answered by combining information from multiple sentences of the paragraph.
The dataset was designed with three key challenges in mind:
* The number of correct answer-options for each question is not pre-specified. This removes the over-reliance on answer-options and forces them to decide on the correctness of each candidate answer independently of others. In other words, the task is not to simply identify the best answer-option, but to evaluate the correctness of each answer-option individually.
* The correct answer(s) is not required to be a span in the text.
* The paragraphs in the dataset have diverse provenance by being extracted from 7 different domains such as news, fiction, historical text etc., and hence are expected to be more diverse in their contents as compared to single-domain datasets.
The entire corpus consists of around 10K questions (including about 6K multiple-sentence questions). The 60% of the data is released as training and development data. The rest of the data is saved for evaluation and every few months a new unseen additional data is included for evaluation to prevent unintentional overfitting over time.

Source: https://cogcomp.seas.upenn.edu/multirc/
Image Source: https://paperswithcode.com/paper/looking-beyond-the-surface-a-challenge-set/

Variants: MultiRC

Associated Benchmarks

This dataset is used in 1 benchmark:

Recent Benchmark Submissions

Task Model Paper Date
Question Answering PaLM 2-L (one-shot) PaLM 2 Technical Report 2023-05-17
Question Answering PaLM 2-S (one-shot) PaLM 2 Technical Report 2023-05-17
Question Answering PaLM 2-M (one-shot) PaLM 2 Technical Report 2023-05-17
Question Answering Bloomberg GPT 50B (1-shot) BloombergGPT: A Large Language Model … 2023-03-30
Question Answering OPT 66B (1-shot) BloombergGPT: A Large Language Model … 2023-03-30
Question Answering GPT-NeoX 20B (1-shot) BloombergGPT: A Large Language Model … 2023-03-30
Question Answering BLOOM 176B (1-shot) BloombergGPT: A Large Language Model … 2023-03-30
Question Answering Hybrid H3 355M (3-shot, logit scoring) Hungry Hungry Hippos: Towards Language … 2022-12-28
Question Answering Hybrid H3 125M (3-shot, logit scoring) Hungry Hungry Hippos: Towards Language … 2022-12-28
Question Answering Hybrid H3 125M (0-shot, logit scoring) Hungry Hungry Hippos: Towards Language … 2022-12-28
Question Answering Hybrid H3 355M (0-shot, logit scoring) Hungry Hungry Hippos: Towards Language … 2022-12-28
Question Answering Vega v2 6B (fine-tuned) Toward Efficient Language Model Pretraining … 2022-12-04
Question Answering Turing NLR v5 XXL 5.4B (fine-tuned) Toward Efficient Language Model Pretraining … 2022-12-04
Question Answering Neo-6B (few-shot) Ask Me Anything: A simple … 2022-10-05
Question Answering Neo-6B (QA) Ask Me Anything: A simple … 2022-10-05
Question Answering Neo-6B (QA + WS) Ask Me Anything: A simple … 2022-10-05
Question Answering AlexaTM 20B AlexaTM 20B: Few-Shot Learning Using … 2022-08-02
Question Answering N-Grammer 343M N-Grammer: Augmenting Transformers with latent … 2022-07-13
Question Answering PaLM 540B (finetuned) PaLM: Scaling Language Modeling with … 2022-04-05
Question Answering ST-MoE-32B 269B (fine-tuned) ST-MoE: Designing Stable and Transferable … 2022-02-17

Research Papers

Recent papers with results on this dataset: