Investigating Multi-Cultural Value Selection Biases in LLMs
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
A study tests how large language models respond when users signal two cultural identities at once, a country persona in the system prompt plus either an explicitly named country or a query in another language.
Score
0.60total under rubric v2.0
- Significance0.55
- Reader interest0.63
- Audience breadth0.59
- Practical relevance0.59
- Substantiated0.74
- Technical depth0.63
- New information0.86
- Counts against: Promotional0.03
- Counts against: Overclaim0.16
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| Significance | 0.64 | 0.40 | 0.62 | 0.55 | 0.23 | +0.25 | +0.138 |
| Reader interest | 0.52 | 0.60 | 0.78 | 0.63 | 0.26 | +0.20 | +0.127 |
| Audience breadth | 0.66 | 0.50 | 0.62 | 0.59 | 0.16 | +0.10 | +0.059 |
| Practical relevance | 0.68 | 0.60 | 0.48 | 0.59 | 0.20 | +0.15 | +0.088 |
| Substantiated | 0.89 | 0.70 | 0.62 | 0.74 | 0.27 | +0.10 | +0.074 |
| Technical depth | 0.71 | 0.60 | 0.58 | 0.63 | 0.13 | +0.10 | +0.063 |
| New information | 0.93 | 0.70 | 0.94 | 0.86 | 0.24 | +0.10 | +0.086 |
| Promotional | 0.05 | 0.02 | 0.02 | 0.03 | 0.03 | −0.15 | −0.004 |
| Overclaim | 0.17 | 0.20 | 0.12 | 0.16 | 0.08 | −0.15 | −0.024 |
| Sum of contributions | +0.605 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.605 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN —
Recorded for context only, checked 24 Oct 2026, 08:54 UTC. It played no part in the ranking.