Anthropocentric bias in language model evaluation
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915668922728448 |
|---|---|
| author | Millière, Raphaël Rathkopf, Charles |
| author_facet | Millière, Raphaël Rathkopf, Charles |
| contents | 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: overlooking how auxiliary factors can impede LLM performance despite competence ("auxiliary oversight"), and dismissing LLM mechanistic strategies that differ from those of humans as not genuinely competent ("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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_03859 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Anthropocentric bias in language model evaluation Millière, Raphaël Rathkopf, Charles Computation and Language 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: overlooking how auxiliary factors can impede LLM performance despite competence ("auxiliary oversight"), and dismissing LLM mechanistic strategies that differ from those of humans as not genuinely competent ("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. |
| title | Anthropocentric bias in language model evaluation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2407.03859 |