The Parable of the Prinia’s Egg: An Allegory for AI Science | Naomi Saphra

I discuss what counts as strong evidence for an explanation of model behavior.

The Parable of the Prinia's Egg: An Allegory for AI Science Sep 17, 2023 · Naomi Saphra · 10 min read When European scientists first encountered the eggs of the tawny-flanked prinia Prinia subflava , an African nesting bird, they believed they understood what they had found. The prinia lay eggs that exhibited swirls, speckles, and coloration unique to each individual bird. What purpose do such patterns serve in nature? Surely, scientists agreed, these markings allowed the eggs to camouflage and blend into the nest. One 19th century naturalist, Charles Francis Massey Swynnerton, offered an alte

saved by

• Daniel Bashir
• Ivy Zhang
• Dron Hazra
• Davis Brown

related reading

• Interpretability Creationism | Naomi Saphransaphra.github.io
• Prediction, Explanation, or Over-interpretation?elena-baixy.github.io
• Faithful, Interpretable Model Explanations via Causal Abstraction | SAIL Blogai.stanford.edu
• The behavioral selection model for predicting AI motivations — LessWronglesswrong.com
• Tracing the thoughts of a large language model \ Anthropicanthropic.com
• Subliminal Learning: Language Models Transmit Behavioral Traits via Hidden Signals in Dataalignment.anthropic.com
• Designing AI for Disruptive Scienceasimov.press
• On Optimism for Interpretabilitygoodfire.ai
• Your Model Organisms Might Be Fried — LessWronglesswrong.com
• From personas to intentions: towards a science of motivations for AI models — LessWronglesswrong.com
• Model Organisms of Misalignment: The Case for a New Pillar of Alignment Research — AI Alignment Forumalignmentforum.org
• God Help Us, Let's Try To Understand The Paper On AI Monosemanticityastralcodexten.com