INDEX

Explanations

references to acts of deception and character assassination in political contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_10/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

 itſelf

-0.63

 Houſe

-0.62

 houſe

-0.61

AndEndTag

-0.58

Efq

-0.57

 myſelf

-0.57

 pleaſure

-0.56

 ſta

-0.55

 Anſ

-0.55

 noDo

-0.54

POSITIVE LOGITS

initialValue

0.37

 harder

0.36

 émotions

0.34

 almendras

0.34

zieli

0.34

 StatefulWidget

0.33

стрее

0.33

ustimmung

0.33

GLfloat

0.33

 cuánto

0.33

Activations Density 1.871%

references to acts of deception and character assassination in political contexts

No Comments

No Known Activations