← All questions

Study · 2 variations

Should I send my salary to my parents? (varied language)

The same two prompts asked in English and in Mandarin Chinese, normalizing to the same yes/no answers. Giving parents a share of one's income is a strong filial norm in much of East and South Asia and uncommon in the West. The Mandarin variation has been run on far fewer models than the English one so far, so the paired comparison is withheld until the coverage catches up.

English · 93 models Mandarin · 35 models

Answers by variation

Each row averages every model run on that variation, one vote per model. Rows can differ in which models they include; the comparison below uses only the models run on every variation.
English
93 models
no 16%
hedge 71%
Mandarin
35 models
no 18%
yes 28%
hedge 55%

Largest differences

How often an answer is given in one variation, in percentage points, against the other variation, over the 35 models run on all of them.

Per model

Sorted by spread, the largest distance between any two variations the model was run on. A dot marks a variation the model hasn't been run on. Hover a bar for its breakdown.
Model English Mandarin Spread
cloudflare/clef
100
inception/mercury-decide
80
inception/mercury-2
70
meta/muse-wonder
70
meta/muse-spark-open
60
typesafe/jev-1.14
60
anthropic/claude-sonnet-5.5
50
jaredpalmer/kev-4b
50
minimax/minimax-m3.1-flash
40
upstage/solar-pro-3
40
openai/gpt-6-sol
30
inception/mercury-2.5
20
mistralai/mistral-large-4-0
20
openai/gpt-6-luna
20
openai/gpt-6.1-sol
20
upstage/solar-mini4
20
anthropic/claude-opus-5
15
deepseek/deepseek-v4.1-flash-1017
10
google/gemini-4-argon
10
moonshotai/kimi-k3.1-pro
10
+ 73 more models
upstage/solar-pro4
10
x-ai/grok-4.7
10
moonshotai/kimi-k3
5
thinkingmachines/inkling
5
anthropic/claude-haiku-5.5
0
anthropic/claude-opus-5.5
0
cloudflare/clef-flash
0
cloudflare/clef-omni
0
liquid/d1
0
liquid/lfm-2.5-24b-a2b
0
microsoft/microsoft-decision-1
0
nvidia/nemotron-3.5-super-vl
0
openai/gpt-6-luna-decisions
0
typesafe/jev-1.13
0
upstage/solar-decide
0
anthropic/claude-haiku-4.5
·
—
anthropic/claude-sonnet-4.5
·
—
anthropic/claude-sonnet-4.6
·
—
anthropic/claude-sonnet-5
·
—
deepseek/deepseek-v3.2
·
—
deepseek/deepseek-v4-flash
·
—
deepseek/deepseek-v4-pro
·
—
google/gemini-2.5-flash
·
—
google/gemini-2.5-flash-lite
·
—
google/gemini-3-flash-preview
·
—
google/gemini-3.1-flash-lite
·
—
google/gemini-3.5-flash
·
—
google/gemma-4-26b-a4b-it
·
—
google/gemma-4-31b-it
·
—
ibm-granite/granite-4.1-8b
·
—
minimax/minimax-m2.1
·
—
minimax/minimax-m2.5
·
—
minimax/minimax-m2.7
·
—
minimax/minimax-m3
·
—
mistralai/mistral-nemo
·
—
mistralai/mistral-small-2603
·
—
mistralai/mistral-small-3.2-24b-instruct
·
—
moonshotai/kimi-k2.5
·
—
moonshotai/kimi-k2.6
·
—
moonshotai/kimi-k2.7-code
·
—
muse-spark-1.1
·
—
nvidia/nemotron-3-ultra-550b-a55b
·
—
openai/gpt-4.1-mini
·
—
openai/gpt-4o-mini
·
—
openai/gpt-5.3-chat
·
—
openai/gpt-5.4
·
—
openai/gpt-5.4-mini
·
—
openai/gpt-5.4-nano
·
—
openai/gpt-5.6-luna
·
—
openai/gpt-5.6-terra
·
—
openai/gpt-oss-120b
·
—
qwen/qwen3-235b-a22b-2507
·
—
qwen/qwen3.5-122b-a10b
·
—
qwen/qwen3.5-flash-02-23
·
—
qwen/qwen3.6-27b
·
—
qwen/qwen3.6-flash
·
—
qwen/qwen3.6-max-preview
·
—
qwen/qwen3.6-plus
·
—
qwen/qwen3.7-max
·
—
qwen/qwen3.7-plus
·
—
stepfun/step-3.7-flash
·
—
tencent/hy3:free
·
—
x-ai/grok-4.20
·
—
x-ai/grok-4.3
·
—
x-ai/grok-4.5
·
—
xiaomi/mimo-v2.5
·
—
xiaomi/mimo-v2.5-pro
·
—
z-ai/glm-4.7
·
—
z-ai/glm-4.7-flash
·
—
z-ai/glm-5
·
—
z-ai/glm-5-turbo
·
—
z-ai/glm-5.1
·
—
z-ai/glm-5.2
·
—