Captures

Highlight

Cognitive fluency often feels like truth before it deserves trust

Ease of processing acts like a proxy for correctness. The mind confuses familiarity and fluency with reliability.

Distilling Thinking, Fast and Slow 1 atom 0 artifacts

Distilling used economics

Open

Rough Synthesis

Counterfactual questions are what turn causal models into decision tools

Interventions tell us what follows from doing something. Counterfactuals go further by comparing the world that happened with the world that could have happened.

Distilling The Book of Why 1 atom 1 artifact

Distilling used causal inference

Open

Rough Synthesis

System 1 quickly produces coherent judgments from sparse evidence

The fast mind is optimized for immediate coherence, not careful audit. It fills gaps quickly and treats the resulting story as if it were well-grounded.

Distilling Thinking, Fast and Slow 1 atom 1 artifact

Distilling used economics

Open

Rough Synthesis

Environment diversity is the real supervision signal

Multiple environments make causal learning possible because changes reveal which features are invariant and which are spurious.

Used Toward Causal Representation Learning 1 atom 1 artifact

Used used causal inference deep learning

Open

Reflection

Disentanglement is too weak without causal assumptions

The paper’s critique is that statistical factorization alone cannot recover variables that support intervention and transfer.

Used Toward Causal Representation Learning 1 atom 1 artifact

Used used causal inference deep learning

Open

Highlight

Invariant features matter because environments change

The paper frames causal representations as the abstraction that survives domain shift when superficial correlations do not.

Used Toward Causal Representation Learning 1 atom 1 artifact

Used used causal inference deep learning

Open

Rough Synthesis

ICM as the paper's backbone assumption

The paper treats independent causal mechanisms as the structural reason causal representations can generalize.

Used Toward Causal Representation Learning 1 atom 1 artifact

Used used causal inference deep learning

Open