INDEX

Explanations

negation phrases or words that convey refusal or denial

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

negation or words starting with ne/ni

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

bootstrapcdn

-0.55

gameserver

-0.50

Datuak

-0.49

ynb

-0.43

 okuyayım

-0.42

Portail

-0.42

BagLayout

-0.42

VersionUID

-0.42

anthin

-0.42

umbersome

-0.41

POSITIVE LOGITS

 Не

0.69

Non

0.68

Non

0.64

Не

0.63

non

0.63

 noDo

0.61

 Nicht

0.60

Not

0.60

 Não

0.60

 не

0.59

Activations Density 0.001%

negation phrases or words that convey refusal or denial

negation or words starting with ne/ni

No Comments

No Known Activations