INDEX

Explanations

usage of the underscore character

oai_token-act-pair · gpt-4o-mini Triggered by @bot

underscore and words starting with unders

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

File

-0.40

It

-0.39

it

-0.39

getB

-0.39

Julian

-0.39

fake

-0.38

test

-0.37

Pir

-0.37

indelijk

-0.37

Login

-0.37

POSITIVE LOGITS

 underscore

1.66

 underscores

1.51

 underscored

1.43

underscore

1.41

 Unders

1.41

 unders

1.30

Unders

1.28

unders

1.17

 underserved

0.66

 undercut

0.65

Activations Density 0.008%

usage of the underscore character

underscore and words starting with unders

No Comments

No Known Activations