Angie Boggust
@angieboggust.bsky.social
MIT PhD candidate in the VIS group working on interpretability and human-AI alignment
Abstraction Alignment works on datasets too!
Medical experts analyzed clinical dataset abstractions, uncovering issues like overuse of unspecified diagnoses.
This mirrors real-world updates to medical abstractions — showing how models can help us rethink human knowledge.
Medical experts analyzed clinical dataset abstractions, uncovering issues like overuse of unspecified diagnoses.
This mirrors real-world updates to medical abstractions — showing how models can help us rethink human knowledge.
April 14, 2025 at 3:48 PM
Abstraction Alignment works on datasets too!
Medical experts analyzed clinical dataset abstractions, uncovering issues like overuse of unspecified diagnoses.
This mirrors real-world updates to medical abstractions — showing how models can help us rethink human knowledge.
Medical experts analyzed clinical dataset abstractions, uncovering issues like overuse of unspecified diagnoses.
This mirrors real-world updates to medical abstractions — showing how models can help us rethink human knowledge.
Abstraction Alignment compares model behavior to human abstractions.
By propagating the model's uncertainty through an abstraction graph, we can see how well it aligns with human knowledge.
E.g., confusing oaks🌳 with palms🌴 is more aligned than confusing oaks🌳 with sharks🦈.
By propagating the model's uncertainty through an abstraction graph, we can see how well it aligns with human knowledge.
E.g., confusing oaks🌳 with palms🌴 is more aligned than confusing oaks🌳 with sharks🦈.
April 14, 2025 at 3:48 PM
Abstraction Alignment compares model behavior to human abstractions.
By propagating the model's uncertainty through an abstraction graph, we can see how well it aligns with human knowledge.
E.g., confusing oaks🌳 with palms🌴 is more aligned than confusing oaks🌳 with sharks🦈.
By propagating the model's uncertainty through an abstraction graph, we can see how well it aligns with human knowledge.
E.g., confusing oaks🌳 with palms🌴 is more aligned than confusing oaks🌳 with sharks🦈.
Interpretability identifies models' learned concepts (wheels 🛞).
But human reasoning is built on abstractions — relationships between concepts that help us generalize (wheels 🛞→ car 🚗).
To measure alignment, we must test if models learn human-like concepts AND abstractions.
But human reasoning is built on abstractions — relationships between concepts that help us generalize (wheels 🛞→ car 🚗).
To measure alignment, we must test if models learn human-like concepts AND abstractions.
April 14, 2025 at 3:48 PM
Interpretability identifies models' learned concepts (wheels 🛞).
But human reasoning is built on abstractions — relationships between concepts that help us generalize (wheels 🛞→ car 🚗).
To measure alignment, we must test if models learn human-like concepts AND abstractions.
But human reasoning is built on abstractions — relationships between concepts that help us generalize (wheels 🛞→ car 🚗).
To measure alignment, we must test if models learn human-like concepts AND abstractions.