INDEX

Explanations

user id, password, authenticated

This neuron activates on occurrences of “user” in identifier-like contexts (e.g. user_id, user_path) in code or schema.

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

Ꮠ

-2.34

饪

-2.17

絎

-2.13

⃞

-2.06

 två

-2.00

 dwóch

-1.96

祧

-1.93

珰

-1.90

 they

-1.89

䡲

-1.87

POSITIVE LOGITS

<i>

2.69

的区别

2.34

</td>

2.03

The

1.98

self

1.95

唛

1.92

1.89

沇

1.88

<b>

1.87

Activations Density 0.051%