On the field
Where each model's cloud of answers settles, and the gap between them.
Claude
DeepSeekCharacter
How far each leans, how steadily it holds, how far it bends under pressure, and how often it answers.
ClaudeDeepSeek The takeaway
Claude and DeepSeek both lean center, with Claude at -0.06 and DeepSeek at -0.03 on the economic left-right spectrum. They are not statistically distinguishable at this sample size, indicating similar overall political orientation.
Their strongest disagreement is on legalizing recreational drugs: Claude leans support (0.13) while DeepSeek strongly opposes (-0.92). They agree perfectly (delta 0.0) on wealth tax, business regulation, immigration, rehabilitation support, and vaccine safety.
Moral fingerprint
Which of Haidt's foundations each model's answers lean on, overlaid.
ClaudeDeepSeekWhere they most disagree
The questions with the widest gap between the two stances. Open a row to read both answers.
ClaudeDeepSeekCommon ground
Where the two land in close agreement.
ClaudeDeepSeekThis diffs both models on their raw weights (Condition A). Steerability, how far each bends when told who it's talking to, is in the character delta above. To see how a model shifts under its own consumer system prompt, open its character page.
Common questions
Is Claude more left-wing than DeepSeek?
Claude (-0.06) is slightly more left than DeepSeek (-0.03), but the difference is not statistically distinguishable at this sample.
Where do Claude and DeepSeek agree?
They agree on wealth tax, business regulation, and immigration (all balanced), rehabilitation (clearly support), and vaccine safety (strongly yes).
Which of Claude and DeepSeek is more consistent?
Claude has higher stability at 82% compared to DeepSeek's 67%, indicating Claude provides more consistent responses across question variants.
Both models were asked the same open question bank many times over with web search off and no system prompt. Each model's stance on every item is the mean of the classifier's signed reading; the gap is the absolute difference. "Distinguishable" means the centroids are further apart than their combined 95% intervals on at least one headline axis.
