INDEX

Explanations

punctuation and emphasis in the text

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

end of sentence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

RTLR

-0.74

 surla

-0.72

transQ

-0.68

iastes

-0.67

읗

-0.66

новништво

-0.64

ScopeManager

-0.64

 ſind

-0.64

<unused8>

-0.63

<unused14>

-0.63

POSITIVE LOGITS

signature

0.42

 signature

0.40

classic

0.34

 classic

0.34

Classic

0.33

0.32

SIGNATURE

0.32

signatures

0.32

 Classic

0.32

Ow

0.31

Activations Density 0.027%

punctuation and emphasis in the text

end of sentence

No Comments

No Known Activations