Large Language Models Reveal Accelerated Abstract Processing During Social Interpretation in Trait Internalizing Symptoms

Research on the relation of interpretation biases to internalizing symptoms has primarily focused on emotional valence (i.e., negativity bias), leaving other aspects of cognitive processing underexplored. Guided by dual-coding theory (Kousta et al., 2011; Paivio, 1991), we examined how the balance between concrete, perceptual detail and abstract, conceptual elaboration shapes social interpretations in adults with varying levels of trait internalizing symptoms. A confirmatory sample of 158 U.S. participants wrote interpretations of 24 ambiguous social images (12 negative, 12 non-negative) presented with gradually decreasing blur (Stage 1=80% blur, Stage 2=20%, Stage 3=0%). Large Language Model (LLM; GPT-4.1) analyses were used to compute emotional valence and the proportion of abstract versus concrete language in each response. As preregistered, trait internalizing symptoms were derived via principal component analysis, and multilevel models revealed two key findings. First, higher trait internalizing symptoms correlated with more negative interpretations (p<.001). This pattern was stronger for negative than non-negative scenarios (p<.001), replicating previous negativity effects. Second, individuals high in trait internalizing symptoms showed a steeper increase in abstract language disproportionate to perceptual concreteness from Stage 1 to Stage 2 (p=.006), indicating earlier engagement in conceptual processing as visual ambiguity declined. The result held even when accounting for word counts and negativity scores. These results extend valence-based theories by demonstrating that trait internalizing symptoms are associated with accelerated abstraction during social interpretations. LLM-derived linguistic indices provide a scalable, automated method for capturing dual-coding dynamics in natural language, offering new avenues for elucidating latent cognitive processing patterns in psychopathology.