INDEX

Explanations

phrases indicating calls to action or invitations to participate

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

call for participation

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

AndEndTag

-0.38

 ")");

-0.37

 garganta

-0.36

 otomatig

-0.35

(()=>{

-0.35

)):

-0.35

++){

-0.34

 ProtoMessage

-0.34

complexContent

-0.34

 Preise

-0.33

POSITIVE LOGITS

GOTREF

0.60

 ſche

0.60

 الحره

0.57

 المعيارى

0.54

principalColumn

0.53

Efq

0.53

 poffe

0.52

 Majefty

0.52

 Anſ

0.51

 occaf

0.50

Activations Density 0.008%

phrases indicating calls to action or invitations to participate

call for participation

No Comments

No Known Activations