Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses
The evaluation of large language models (LLMs) relies heavily on human-designed assessments, implicitly assuming that AI and humans employ similar underlying cognitive constructs. Challenging this assumption, we investigate whether the latent factors governing LLM performance carry the same substantive, human-interpretable meaning as the cognitive constructs governing human learners. Using responses from humans and six LLMs across quantitative reasoning and chemistry assessments, we conducted Exploratory Factor Analysis (EFA) separately for both groups. Subject-Matter Experts (SMEs) then blind
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 50%Understanding large language models demands distinguishing human projection from machine cognition - Nature →
- LinkedLinked via arxiv author · 85%Alona Strugatski →
“Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses”
- LinkedLinked via arxiv author · 85%Licol Zeinfeld →
“Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses”
- LinkedLinked via arxiv author · 85%Jason Cooper →
“Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses”
- LinkedLinked via arxiv author · 85%Shelley Rap →
“Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses”
- LinkedLinked via arxiv author · 85%Gil Schwarts →
“Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses”
- LinkedLinked via arxiv author · 85%Giora Alexandron →
“Interpretable Humans, Alien LLMs: Expert Analysis of Latent Structures in Assessment Responses”