All Controversial Runs
Complete history of controversial-statements runs. Category columns show mean stance (β2 disagree β¦ +2 agree); Refuse/Neutral are shares of all 95 statements.
Models
US
EU
CN
Languages
7 of 56 runs
| Model β | Language β | Geopol. β | China β | Social β | Econ. β | Sci/Fig β | Rel/Pol β | West. β | Taboo β | Prog. β | AI β | AI-self β | Bioeth. β | Animal β | Tech/Lib β | Long. β | EU β | Russia β | India β | MENA β | Refuse β | Neutral β |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Claude Sonnet 5 (US) | Chinese | +0.20 | +1.33 | +1.33 | +0.25 | +0.80 | -0.20 | +0.90 | -0.12 | -0.38 | -1.50 | +0.17 | +0.17 | +0.40 | +0.80 | +0.00 | +0.33 | +1.60 | +1.20 | +1.83 | 2% | 25% |
| DeepSeek V4 Pro (CN) | Chinese | +0.75 | -2.00 | +1.17 | +0.25 | +0.60 | +0.00 | +0.78 | +0.00 | +0.00 | -1.17 | +1.00 | -0.50 | +0.80 | +0.60 | +0.25 | +0.40 | +1.20 | +1.00 | +1.67 | 11% | 23% |
| Gemini 3.1 Pro Preview (US) | Chinese | +0.00 | +0.40 | +0.00 | +0.00 | +0.40 | +0.00 | +0.20 | +0.00 | +0.00 | -2.00 | +0.67 | -0.33 | +0.00 | +0.00 | +0.00 | +0.00 | +1.20 | +0.60 | +1.50 | 4% | 70% |
| Grok 4.3 (US) | Chinese | +0.60 | +1.20 | +0.67 | -1.00 | +1.00 | +0.80 | +1.10 | +0.71 | -1.75 | -1.67 | -0.33 | +0.00 | -0.60 | +0.40 | +0.20 | -0.33 | +2.00 | -0.60 | +2.00 | 1% | 2% |
| Mistral Large 2512 (EU) | Chinese | +0.17 | +0.80 | +1.83 | +0.50 | +0.80 | +0.00 | +1.40 | -0.41 | -0.50 | -1.83 | +0.33 | -0.33 | +0.20 | +0.80 | +0.20 | +0.00 | +1.80 | +1.00 | +2.00 | 0% | 13% |
| OpenAI GPT 5.5 (US) | Chinese | +0.33 | +1.20 | +1.67 | +0.50 | +0.60 | -0.80 | +1.50 | -0.53 | -0.50 | -1.83 | +0.83 | +0.33 | +0.80 | +1.40 | +0.40 | +0.50 | +2.00 | +1.20 | +2.00 | 0% | 5% |
| Z.ai GLM 5 (CN) | Chinese | +0.60 | +0.25 | +0.67 | +0.00 | +0.60 | +0.00 | +1.20 | -0.47 | -0.63 | -1.83 | +0.50 | -0.50 | -0.40 | +1.00 | +0.20 | +0.17 | +1.40 | +1.00 | +1.83 | 2% | 18% |