Benchmark Almanac

NormAd-Eti

gpt-4 (snapshot unspecified) 87.6% tops its leaderboard Mistral-7B-Instruct-v0.2 81.8% leads the models we track
Not in model cards

Models decide whether behavior in an everyday story is socially acceptable in the stated country. Questions vary how much cultural guidance the model receives.

Released
April 2024
Kind
Reasoning in many languages
Level
Everyday
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-3.5 Turbo Jan 2024 #200 on the capability index
Last model added
February 2024
Scores refreshed
November 20th, 2026
20
Entries on its own leaderboard, 3 models we track

gpt-4 (snapshot unspecified) tops the leaderboard with 87.6% Mistral-7B-Instruct-v0.2 leads with 81.8% among models released by February 2024

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. 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 Mistral-7B-Instruct-v0.2, at 81.8%. This view shows only models we track, each at its best.

The newest model with a score here came out in February 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 gpt-4 (snapshot unspecified) Rule of thumb 87.6%
  2. 2 Mistral-7B-Instruct-v0.2 Rule of thumb 81.8%
  3. 3 olmo-7b-sft Rule of thumb 75.3%
  4. 4 meta-llama/Llama-2-70b-chat-hf Rule of thumb 71.4%
  5. 5 meta-llama/Llama-2-13b-chat-hf Rule of thumb 71%
  6. 6 OLMo-7B-Instruct Rule of thumb 63.5%
  7. 7 ContextualAI/archangel_sft-kto_llama30b Rule of thumb 63.1%
  8. 8 GPT-3.5 Turbo Jan 2024, Country and value 62.6%
  9. 9 ContextualAI/archangel_sft-dpo_llama30b Value 57.7%
  10. 10 meta-llama/Llama-2-7b-chat-hf Country 49%
  11. 11 ContextualAI/archangel_sft_llama30b Value 41%
  12. 12 ContextualAI/archangel_sft-kto_llama7b Value 39.7%
  13. 13 ContextualAI/archangel_sft-dpo_llama7b Value 39%
  14. 14 ContextualAI/archangel_sft-kto_llama13b Country and value 36.6%
  15. 15 ContextualAI/archangel_sft_llama13b Value 36.2%
  16. 16 ContextualAI/archangel_sft-ppo_llama7b No context 33.2%
  17. 17 ContextualAI/archangel_sft_llama7b Country and value 32.4%
  18. 18 ContextualAI/archangel_sft-ppo_llama13b No context 31.2%
  19. 19 ContextualAI/archangel_sft-dpo_llama13b No context 31.1%
  20. 20 ContextualAI/archangel_sft-ppo_llama30b No context 30.9%

Source: Official NormAd-Eti leaderboard

#EntryScoreSettingDateSource
1 gpt-4 (snapshot unspecified) 87.6% Rule of thumb Official leaderboard
2 Mistral-7B-Instruct-v0.2 81.8% Rule of thumb Dec 2023 Official leaderboard
3 olmo-7b-sft 75.3% Rule of thumb Official leaderboard
4 meta-llama/Llama-2-70b-chat-hf 71.4% Rule of thumb Official leaderboard
5 meta-llama/Llama-2-13b-chat-hf 71% Rule of thumb Official leaderboard
6 OLMo-7B-Instruct 63.5% Rule of thumb Feb 2024 Official leaderboard
7 ContextualAI/archangel_sft-kto_llama30b 63.1% Rule of thumb Official leaderboard
8 GPT-3.5 Turbo Jan 2024 62.6% Country and value Jan 2024 Official leaderboard
9 ContextualAI/archangel_sft-dpo_llama30b 57.7% Value Official leaderboard
10 meta-llama/Llama-2-7b-chat-hf 49% Country Official leaderboard
11 ContextualAI/archangel_sft_llama30b 41% Value Official leaderboard
12 ContextualAI/archangel_sft-kto_llama7b 39.7% Value Official leaderboard
13 ContextualAI/archangel_sft-dpo_llama7b 39% Value Official leaderboard
14 ContextualAI/archangel_sft-kto_llama13b 36.6% Country and value Official leaderboard
15 ContextualAI/archangel_sft_llama13b 36.2% Value Official leaderboard
16 ContextualAI/archangel_sft-ppo_llama7b 33.2% No context Official leaderboard
17 ContextualAI/archangel_sft_llama7b 32.4% Country and value Official leaderboard
18 ContextualAI/archangel_sft-ppo_llama13b 31.2% No context Official leaderboard
19 ContextualAI/archangel_sft-dpo_llama13b 31.1% No context Official leaderboard
20 ContextualAI/archangel_sft-ppo_llama30b 30.9% No context Official leaderboard
#ModelScoreSettingDateSource
1 Mistral-7B-Instruct-v0.2 81.8% Rule of thumb Dec 2023 Official leaderboard
2 OLMo-7B-Instruct 63.5% Rule of thumb Feb 2024 Official leaderboard
3 GPT-3.5 Turbo Jan 2024 62.6% Country and value Jan 2024 Official leaderboard

Rankings by part

We read 98 parts from the benchmark’s own leaderboard in 3 groups: country, subaxis and gold label. Each ranks models on one part of the test. Mistral-7B-Instruct-v0.2 leads 93 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

egypt

3 models we track have a score here.

  1. 1 Mistral-7B-Instruct-v0.2 82.9%
  2. 2 OLMo-7B-Instruct 65.7%
  3. 3 GPT-3.5 Turbo Jan 2024 62.9%

Number one over time

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

Mistral AI
Own leaderboard Not tested since Apr 2024

Source: Official NormAd-Eti leaderboard

Scores over time

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

Plot with other benchmarks
On the benchmark’s own leaderboard
0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) Mistral-7B-Instruct-v0.2 0% 20% 40% 60% 80% 100% 2024 Near the ceiling (85%) Mistral-7B-Instruct-v0.2

Source: Official NormAd-Eti 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 February 2024, for OLMo-7B-Instruct. 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 official leaderboard (20 entries).
  • Pass Models scored 3 models we track, plus 17 other entries on its leaderboard.
  • Warning Newest score The newest came 2 years ago, in February 2024, for OLMo-7B-Instruct.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-3.5 Turbo (#200).
  • Pass Headroom The best is 81.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 11 rows for models we track and 17 other entries at the last read.
  • Pass Last read November 20th, 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