Benchmark Almanac

JCoLA

Waseda RoBERTa large (s128) 66.5% tops its leaderboard GPT-4 Jun 2023 54.5% leads the models we track
Not in model cards

Models decide whether Japanese sentences are grammatically acceptable. Linguists supply judgments for sentences from textbooks and the research literature.

The authors report separate results for in-domain and out-of-domain sentences.

Made by
Taiga Someya and colleagues
Released
February 2022
Kind
Reasoning in many languages
Level
Everyday
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
Japanese
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-4 Jun 2023 #186 on the capability index
Last model added
January 2024
Scores refreshed
November 21st, 2026
11
Entries in its paper, 2 models we track

Waseda RoBERTa large (s128) tops the leaderboard with 66.5% GPT-4 leads with 54.5%

Best score per entry from the benchmark's paper, read on November 21st, 2026. A paper’s table doesn’t add newer models. The paper reports no overall score, so this is the plain mean of its in-domain and out-of-domain accuracy columns. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is GPT-4, at 54.5%. This view shows only models we track, each at its best.

  1. 1 Waseda RoBERTa large (s128) 66.5%
  2. 2 Waseda RoBERTa large (s512) 63.8%
  3. 3 Waseda RoBERTa base 63.5%
  4. 4 NICT BERT base 60.9%
  5. 5 Tohoku BERT base 59.8%
  6. 6 Tohoku BERT large 59.4%
  7. 7 GPT-4 Jun 2023 54.5%
  8. 8 XLM RoBERTa large 53.8%
  9. 9 Tohoku BERT base (char) 51.9%
  10. 10 XLM RoBERTa base 48.8%
  11. 11 GPT-3.5 Turbo Jan 2024 40.5%

Source: The JCoLA paper

Source: The JCoLA paper

#EntryScoreSettingDateSource
1 Waseda RoBERTa large (s128) 66.5% Official leaderboard
2 Waseda RoBERTa large (s512) 63.8% Official leaderboard
3 Waseda RoBERTa base 63.5% Official leaderboard
4 NICT BERT base 60.9% Official leaderboard
5 Tohoku BERT base 59.8% Official leaderboard
6 Tohoku BERT large 59.4% Official leaderboard
7 GPT-4 Jun 2023 54.5% Jun 2023 Official leaderboard
8 XLM RoBERTa large 53.8% Official leaderboard
9 Tohoku BERT base (char) 51.9% Official leaderboard
10 XLM RoBERTa base 48.8% Official leaderboard
11 GPT-3.5 Turbo Jan 2024 40.5% Jan 2024 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-4 Jun 2023 54.5% Jun 2023 Official leaderboard
2 GPT-3.5 Turbo Jan 2024 40.5% Jan 2024 Official leaderboard

Scores over time

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

In the benchmark’s paper
0% 20% 40% 60% 80% 100% 2023 2024 Near the ceiling (85%) GPT-4 0% 20% 40% 60% 80% 100% 2023 2024 Near the ceiling (85%) GPT-4

Source: Official JCoLA leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Multilingual

All 545 in Multilingual
  • MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
  • WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET. K-EXAONE 90.5%
  • MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
  • CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories. Solar Open 2 90.7%
  • Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
  • MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in January 2024, for GPT-3.5 Turbo. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (11 entries).
  • Pass Models scored 2 models we track, plus 9 other entries on its leaderboard.
  • Warning Newest score The newest came 2 years ago, in January 2024, for GPT-3.5 Turbo.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4 (#186).
  • Pass Headroom The best is 54.5%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on aclanthology.org: 2 rows for models we track and 9 other entries at the last read.
  • Pass Last read November 21st, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check