INDEX

Explanations

names and specific identifiers related to individuals and entities

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

ind/ink + er/ver

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

={({

-0.58

ztály

-0.56

didSet

-0.55

epa

-0.53

Eq

-0.53

sApp

-0.52

者

-0.50

$\#

-0.49

ábado

-0.49

 Koval

-0.49

POSITIVE LOGITS

int

0.64

inton

0.63

 Wink

0.63

ind

0.61

ink

0.61

inn

0.61

into

0.60

inta

0.60

inde

0.59

inthe

0.59

Activations Density 0.052%

names and specific identifiers related to individuals and entities

ind/ink + er/ver

No Comments

No Known Activations