RTE

PaLM 2-L 79.3%

T5-11B 92.8% leads

T5-11B 92.8% leads since October 2019

Dropped

Labs’ own numbers All independent runs Include claimed scores

Models decide whether a premise sentence entails a hypothesis, using data from the PASCAL RTE challenges. Scored by accuracy.

Made by: PASCAL RTE Challenge organizers, packaged in GLUE
Released: April 2018
Kind: Language reasoning
Level: Everyday
Scored: Checked automatically
Answers: Multiple choice
Tasks: All public

Strongest model tested: Claude 2 #191 on the capability index
Last in a launch post: May 2023, dropped after it saturated
Last model added: July 2023
Scores refreshed: November 12th, 2026

Model scored in lab cards

Launch post lists it, from 1 lab

PaLM 2-L leads with 79.3%

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.

T5-11B leads with 92.8% 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.

T5-11B leads with 92.8%, 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

79.3 PaLM 2-L

1 PaLM 2-L 79.3%

Source: Labs’ model cards

1 T5-11B 92.8%
2 Claude 2 54.1%

Source: Official RTE leaderboard

1 T5-11B 92.8%
2 PaLM 2-L 79.3%
3 Claude 2 54.1%

Sources: Official RTE leaderboard and labs’ model cards

#	Model	Score	Setting	Date	Source
1	PaLM 2-L	79.3%		May 2023	Lab model card

#	Model	Score	Setting	Date	Source
1	T5-11B	92.8%		Oct 2019	Official leaderboard
2	Claude 2	54.1%		Jul 2023	Official leaderboard

#	Model	Score	Date	Source
1	T5-11B	92.8%	Oct 2019	Official leaderboard
2	PaLM 2-L	79.3%	May 2023	Lab model card
3	Claude 2	54.1%	Jul 2023	Official leaderboard

The benchmark's own leaderboard has 263 entries, 2 of them models we track. Switch the scores above to rank them together.

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

Sources: Official RTE 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.

Google DeepMind

Date	Launch post	Score	As the post writes it
May 10, 2023	PaLM 2-L	79.3	RTE, 1-shot

PaLM 2-L leads with 79.3%

T5-11B leads with 92.8% across all independent runs

T5-11B leads with 92.8%, claims included

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

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

More in Knowledge