INDEX

Explanations

references to evil characters and their interactions within a narrative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

цездатний

-0.54

Personensuche

-0.50

RTEX

-0.46

 milagro

-0.43

__':

-0.43

 VIDA

-0.41

peo

-0.40

 Dalio

-0.40

henge

-0.40

 Reſ

-0.40

POSITIVE LOGITS

 nahilalakip

0.53

istoitu

0.53

 terrorists

0.50

 evil

0.49

 villain

0.48

 Satan

0.46

 monsters

0.45

 monster

0.44

 monstre

0.43

 criminals

0.42

Activations Density 0.485%

references to evil characters and their interactions within a narrative context

No Comments

No Known Activations