INDEX

Explanations

references to privacy policy changes

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

policy changes

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

eroon

-0.42

yyr

-0.41

koff

-0.39

 choke

-0.39

TokenNameLPAREN

-0.39

hoff

-0.39

edig

-0.38

筋

-0.38

agit

-0.38

whelmed

-0.38

POSITIVE LOGITS

 AssemblyCulture

0.50

":[{

0.47

styleType

0.45

 válto

0.41

 changes

0.41

曖昧さ回避

0.41

 AssemblyCompany

0.41

 perubahan

0.40

日閲覧

0.40

 comuniques

0.38

Activations Density 0.005%

references to privacy policy changes

policy changes

No Comments

No Known Activations