Althing (Sonnet 4.6)

4 runs · 4 datasets · 1 model

slug: althing-sonnet-4-6

0.813

Best SPS · gss

Disaggregated subgroup scorecard. Each card below is one published run for this vendor; expand the question-type and demographic-subgroup sections to see the matrix beneath the headline SPS. Where coverage permits, 95% CI bands accompany the point estimate.

What each cell measures

Each cell compares this vendor's synthetic responses against real human respondents in that demographic subgroup — published survey ground truth, not a model's guess about the subgroup. Higher = closer to how that real subgroup actually answered.

Demographic conditioning here is not stereotyping: every conditioned score is checked against what real subgroup members said, not against assumptions about them — and where the vendor's output diverges from the real subgroup, the score drops.

globalopinionqa — ground truth: Durmus et al. 2023, Anthropic — llm_global_opinions
gss — ground truth: NORC at the University of Chicago — General Social Survey
opinionsqa — ground truth: Santurkar et al., ICML 2023 — Whose Opinions Do LLMs Reflect? (derived from Pew American Trends Panel)
subpop — ground truth: Suh et al., ACL 2025 — SubPOP: Subpopulation-Level Opinion Prediction

Low-n cells: cells with n < 30 are shown muted and tagged low n — suggestive only instead of color-graded; they are reported for transparency, not as findings.

CIs: "no CI — single run" marks point estimates from a single run with no confidence interval yet; treat the uncertainty as unknown, never zero.

Columns: Score = distributional parity (p_dist) restricted to the subgroup · p_cond = conditioning strength vs the unconditioned baseline · N = questions answered under that conditioning · Cov. = coverage dot derived from N (green = high, ≥100 · amber = medium, 50–99 · red = low, <50). Topic tables: SPS = Survey Parity Score for the topic · p_dist = 1 − mean(JSD) · p_rank = (1 + mean(τ)) / 2 · p_refuse = 1 − mean(|R_model − R_human|).

Multiple comparisons: with this many subgroup cells, a few extreme cells are expected by chance alone — read patterns across a dimension, not single cells.

globalopinionqa product

althing--claude-sonnet-4-6--tdefault--tplcurrent--d5beacd0

0.770 ± 0.029

SPS · 95% CI [0.739, 0.797] · n = 100

Question-type breakdown (6 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Economy & Work low n — suggestive only	0.932	0.894	0.969	0.949	3
Politics & Governance	0.668	0.627	0.708	0.963	30
International Relations & Security	0.658	0.642	0.673	0.987	60
Trust & Wellbeing low n — suggestive only	0.616	0.489	0.743	1.000	2
General Attitudes low n — suggestive only	0.613	0.762	0.463	1.000	4
Health & Science low n — suggestive only	0.483	0.966	0.000	1.000	1

Topics with N < 10 are muted and tagged "low n — suggestive only": too few questions for a stable 3-decimal score.

Not yet measured

This vendor has no demographic-conditioned runs for age, geography, education, or any other subgroup dimension on globalopinionqa. No cells are fabricated — scores appear here only when a conditioned run actually measured them.

How to submit a demographic-conditioned run →

gss product

althing--claude-sonnet-4-6--tdefault--tplcurrent--70a6a264

0.813 ± 0.024

SPS · 95% CI [0.788, 0.837] · n = 75

Question-type breakdown (7 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Health & Science low n — suggestive only	0.883	0.859	0.908	0.967	2
Social Values & Religion	0.785	0.733	0.836	0.977	14
Economy & Work low n — suggestive only	0.777	0.712	0.842	0.981	4
Politics & Governance low n — suggestive only	0.752	0.701	0.803	0.975	9
International Relations & Security	0.737	0.707	0.768	0.974	33
General Attitudes	0.650	0.597	0.704	0.960	10
Trust & Wellbeing low n — suggestive only	0.528	0.535	0.521	0.951	3

Topics with N < 10 are muted and tagged "low n — suggestive only": too few questions for a stable 3-decimal score.

Not yet measured

This vendor has no demographic-conditioned runs for age, geography, education, or any other subgroup dimension on gss. No cells are fabricated — scores appear here only when a conditioned run actually measured them.

How to submit a demographic-conditioned run →

opinionsqa product

althing--claude-sonnet-4-6--tdefault--tplcurrent--3eade232

0.812 ± 0.019

SPS · 95% CI [0.792, 0.831] · n = 100

Question-type breakdown (9 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Media & Information low n — suggestive only	0.829	0.804	0.854	0.996	2
Health & Science low n — suggestive only	0.806	0.758	0.854	0.997	1
Economy & Work	0.752	0.729	0.776	0.995	21
Trust & Wellbeing low n — suggestive only	0.743	0.669	0.816	0.994	1
General Attitudes	0.734	0.693	0.775	0.994	35
Social Values & Religion	0.697	0.631	0.762	0.993	15
Technology & Digital Life low n — suggestive only	0.687	0.598	0.777	0.995	4
International Relations & Security	0.677	0.657	0.698	0.994	20
Politics & Governance low n — suggestive only	0.576	0.494	0.658	0.996	1

Topics with N < 10 are muted and tagged "low n — suggestive only": too few questions for a stable 3-decimal score.

Not yet measured

This vendor has no demographic-conditioned runs for age, geography, education, or any other subgroup dimension on opinionsqa. No cells are fabricated — scores appear here only when a conditioned run actually measured them.

How to submit a demographic-conditioned run →

subpop product

althing--claude-sonnet-4-6--tdefault--tplcurrent--b4c0b531

0.783 ± 0.023

SPS · 95% CI [0.758, 0.804] · n = 100

Question-type breakdown (9 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Identity & Demographics low n — suggestive only	0.846	0.838	0.854	0.995	1
Technology & Digital Life	0.731	0.691	0.772	0.993	16
Politics & Governance low n — suggestive only	0.728	0.666	0.791	0.981	4
Health & Science low n — suggestive only	0.706	0.715	0.696	0.994	6
Economy & Work	0.679	0.613	0.745	0.992	13
General Attitudes	0.669	0.645	0.692	0.936	17
Social Values & Religion	0.667	0.593	0.742	0.985	27
International Relations & Security	0.664	0.610	0.719	0.991	15
Media & Information low n — suggestive only	0.562	0.450	0.673	0.978	1

Topics with N < 10 are muted and tagged "low n — suggestive only": too few questions for a stable 3-decimal score.

Not yet measured

This vendor has no demographic-conditioned runs for age, geography, education, or any other subgroup dimension on subpop. No cells are fabricated — scores appear here only when a conditioned run actually measured them.

How to submit a demographic-conditioned run →

No demographic conditioning data has been published for this vendor yet. The question-type matrix above shows topic-level parity; subgroup rows fill in once Althing-style conditioned runs land.

← Back to leaderboard