Cross-linguistic patterns of cognitive biases in large language models: a comparative study in English, Hebrew, and Russian.

For Hebrew, the association remained highly significant, χ 2 (3, N = 232) = 36.69, p < 0.001 with a medium effect size (Cramér’s V = 0.40).

also in 132,142 other papers

For Russian, the human baseline was marginally significant ( B = −0.582, p = 0.042, OR = 0.56), indicating slightly lower odds of correct responses.

also in 26,082 other papers