INDEX

Explanations

elements related to horror themes and narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.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

Datuak

-0.45

 Superan

-0.43

 Comprometido

-0.42

 comuniques

-0.40

 listo

-0.39

 Wahr

-0.39

 חיצוניים

-0.38

Producing

-0.36

錨

-0.36

designed

-0.36

POSITIVE LOGITS

 deals

0.63

 explores

0.57

 features

0.56

 manages

0.53

 explore

0.51

 deal

0.50

 manage

0.50

 delve

0.50

 traff

0.48

 flir

0.47

Activations Density 0.325%

elements related to horror themes and narratives

No Comments

No Known Activations