INDEX

Explanations

references to the name "Joseph."

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

<bos>

-0.57

kháu

-0.44

 promp

-0.40

 évêque

-0.40

complexContent

-0.40

 allegation

-0.39

momix

-0.38

 concurso

-0.36

 commandement

-0.35

 komis

-0.35

POSITIVE LOGITS

ValueStyle

0.75

 Chwiliwch

0.72

 Joseph

0.67

frac

0.61

sum

0.59

Joseph

0.59

 JOSEPH

0.54

JOSEPH

0.52

religious

0.51

 joseph

0.51

Activations Density 0.131%

references to the name "Joseph."

No Comments

No Known Activations