INDEX

Explanations

self-consistent, recursion, points

The neuron activates on technical phrases naming iterative or fixed‐point solution methods (e.g. “backward SDE,” “fixed point approach,” “iterative method,” “self‐consistent”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 then

-1.56

its

-1.44

 entonces

-1.41

 wich

-1.37

 メニュー

-1.34

部分が

-1.31

了一個

-1.27

 another

-1.26

 know

-1.23

But

-1.22

POSITIVE LOGITS

})$.

1.43

اد

1.42

我们的

1.41

före

1.38

 Initial

1.38

,”

1.37

他们

1.37

我们

1.36

}_{

1.34

}$

1.33

Activations Density 0.056%