MultiRC

PaLM 2-L 88.2%

PaLM (540B) 88.7% leads since April 2022

Dropped

Labs’ own numbers All independent runs Include claimed scores

Models pick all correct answers to questions that need reasoning over several sentences of a paragraph. Scored by F1 over answer options.

Saturated. Also part of SuperGLUE.

Made by: University of Pennsylvania (CogComp)
Released: June 2018
Kind: Reading comprehension
Level: School
Scored: Checked automatically
Answers: Multiple choice
Tasks: Public, plus a private test set

Strongest model tested: PaLM 2-L #195 on the capability index
Last in a launch post: May 2023, dropped after it saturated
Last model added: May 2023
Scores refreshed: November 12th, 2026

In the authors’ words

We present a reading comprehension challenge in which questions can only be answered by taking into account information from multiple sentences. We solicit and verify questions and answers for this challenge through a 4-step crowdsourcing experiment. Our challenge dataset contains 6,500+ questions for 1000+ paragraphs across 7 different domains (elementary school science, news, travel guides, fiction stories, etc) bringing in linguistic diversity to the texts and to the questions wordings. On a subset of our dataset, we found human solvers to achieve an F1-score of 88.1%. We analyze a range of baselines, including a recent state-of-art reading comprehension system, and demonstrate the difficulty of this challenge, despite a high human performance. The dataset is the first to study multi-sentence inference at scale, with an open-ended set of question types that requires reasoning skills.

Abstract of Looking Beyond the Surface: A Challenge Set for Reading Comprehension over Multiple Sentences (ACL N18-1023, CC BY 4.0)

Model scored in lab cards

Launch post lists it, from 1 lab

PaLM 2-L leads with 88.2%

Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.

PaLM (540B) leads with 88.7% across all independent runs

Best score per model across the official leaderboard, Epoch AI’s own runs and any board Epoch AI copied. Each bar and row names its source.

PaLM (540B) leads with 88.7%, claims included

Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model.

Bars Table

88.2 PaLM 2-L

1 PaLM 2-L 88.2%

Source: Labs’ model cards

Source: Official MultiRC leaderboard

Sources: Official MultiRC leaderboard and labs’ model cards

#	Model	Score	Setting	Date	Source
1	PaLM 2-L	88.2%		May 2023	Lab model card

#	Model	Score	Setting	Date	Source
1	PaLM (540B)	88.7%	F1a	Apr 2022	Official leaderboard
2	T5-11B	88.1%	F1a	Oct 2019	Official leaderboard
3	GPT-3 175B (davinci)	75.4%	F1a	May 2020	Official leaderboard

#	Model	Score	Setting	Date	Source
1	PaLM (540B)	88.7%	F1a	Apr 2022	Official leaderboard
2	PaLM 2-L	88.2%		May 2023	Lab model card
3	T5-11B	88.1%	F1a	Oct 2019	Official leaderboard
4	GPT-3 175B (davinci)	75.4%	F1a	May 2020	Official leaderboard

The benchmark's own leaderboard has 25 entries, 3 of them models we track. Switch the scores above to rank them together.

Number one over time Number one over time Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Google DeepMind

Own leaderboard Passed 85% in Oct 2019, by T5-11B

PaLM (540B) 88.7%

Source: Official MultiRC leaderboard

T5-11B on the official leaderboard passed 85% in October 2019

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Reported by the lab in its model card Best so far

Sources: Official MultiRC leaderboard and labs’ model cards

Labs dropped it from launch posts once the best score passed 85%

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.