guest ::
| Home > Publications database > Anthropocentric bias in language model evaluation |
| Home > Publications database > Anthropocentric bias in language model evaluation |
| Talk (non-conference) (Other) | FZJ-2025-05123 |
2025
Abstract: Evaluating the cognitive capacities of large language models (LLMs) requires overcoming not only anthropomorphic but also anthropocentric biases. This article identifies two types of anthropocentric bias that have been neglected: (i) overlooking how auxiliary factors can impede LLM performance despite competence, which we call auxiliary oversight, and (ii) dismissing LLM mechanistic strategies that differ from those of humans as not genuinely competent, which we call mechanistic chauvinism. Mitigating these biases necessitates an empirically-driven, iterative approach to mapping cognitive tasks to LLM-specific capacities and mechanisms, which can be done by supplementing carefully designed behavioral experiments with mechanistic studies.Paper coauthored with Raphaël Millière.
|
The record appears in these collections: |