INDEX

Explanations

instances of commas in context with varied activation values indicating their importance in the text

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

special characters and symbols

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

-0.50

.”

-0.44

."

-0.42

AndEndTag

-0.41

StoreMessageInfo

-0.39

 Continue

-0.38

 cercando

-0.38

<h2>

-0.38

 clever

-0.38

 heavily

-0.37

POSITIVE LOGITS

+,

0.50

 €,

0.48

uxxxx

0.47

ंदीखरीदारी

0.46

 CanadaChoose

0.45

&,

0.45

Autoritní

0.44

၊

0.44

 informée

0.43

Schweiz

0.43

Activations Density 0.210%

instances of commas in context with varied activation values indicating their importance in the text

special characters and symbols

No Comments

No Known Activations