Althing (Gemini 3.5 Flash-Lite)

4 runs · 4 datasets · 1 model

slug: althing-gemini-3-5-flash-lite

0.839

Best SPS · opinionsqa

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--gemini-3.5-flash-lite--tdefault--tplcurrent--09cfc7e5

0.779 ± 0.030

SPS · 95% CI [0.750, 0.810] · n = 100

Question-type breakdown (6 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Economy & Work low n — suggestive only	0.941	0.912	0.969	0.949	3
General Attitudes low n — suggestive only	0.727	0.741	0.713	1.000	4
Politics & Governance	0.679	0.652	0.705	0.949	30
Trust & Wellbeing low n — suggestive only	0.669	0.588	0.750	1.000	2
International Relations & Security	0.668	0.655	0.682	0.987	60
Health & Science low n — suggestive only	0.492	0.984	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--gemini-3.5-flash-lite--tdefault--tplcurrent--dc95a0a6

0.815 ± 0.024

SPS · 95% CI [0.790, 0.838] · 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	0.807	0.763	0.851	0.977	14
Politics & Governance low n — suggestive only	0.763	0.712	0.814	0.975	9
International Relations & Security	0.724	0.689	0.759	0.974	33
Economy & Work low n — suggestive only	0.719	0.698	0.740	0.981	4
General Attitudes	0.671	0.608	0.734	0.960	10
Trust & Wellbeing low n — suggestive only	0.595	0.590	0.600	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--gemini-3.5-flash-lite--tdefault--tplcurrent--20a55110

0.839 ± 0.018

SPS · 95% CI [0.821, 0.857] · n = 100

Question-type breakdown (9 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Politics & Governance low n — suggestive only	0.862	0.805	0.918	0.996	1
Health & Science low n — suggestive only	0.806	0.758	0.854	0.997	1
Technology & Digital Life low n — suggestive only	0.791	0.771	0.812	0.995	4
Social Values & Religion	0.791	0.762	0.820	0.993	15
Economy & Work	0.783	0.744	0.821	0.995	21
International Relations & Security	0.763	0.742	0.783	0.994	20
Trust & Wellbeing low n — suggestive only	0.743	0.669	0.816	0.994	1
General Attitudes	0.733	0.707	0.760	0.994	35
Media & Information low n — suggestive only	0.701	0.744	0.658	0.996	2

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--gemini-3.5-flash-lite--tdefault--tplcurrent--8892e804

0.811 ± 0.020

SPS · 95% CI [0.791, 0.831] · n = 100

Question-type breakdown (9 topics)

Topic	SPS	p_dist	p_rank	p_refuse	N
Identity & Demographics low n — suggestive only	0.946	0.936	0.956	0.995	1
Health & Science low n — suggestive only	0.811	0.799	0.824	0.994	6
Politics & Governance low n — suggestive only	0.797	0.736	0.857	0.981	4
Technology & Digital Life	0.753	0.715	0.790	0.993	16
General Attitudes	0.741	0.713	0.769	0.936	17
Economy & Work	0.710	0.659	0.761	0.992	13
Social Values & Religion	0.698	0.640	0.755	0.985	27
International Relations & Security	0.688	0.651	0.725	0.991	15
Media & Information low n — suggestive only	0.567	0.505	0.629	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