INDEX

Explanations

references to names or entities in a historical or literary context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_7/width_16k/average_l0_92

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 Majefty

-0.57

IPCC

-0.47

uſ

-0.45

 ProtoMessage

-0.45

 ſte

-0.45

 ſche

-0.44

 itſelf

-0.42

 tranſ

-0.42

GPS

-0.42

 lyre

-0.41

POSITIVE LOGITS

 defaultstate

0.63

IsContent

0.60

<bos>

0.56

 <=",

0.52

تقاوى

0.52

tagHelperRunner

0.51

postIndex

0.50

IsMutable

0.48

tonsoft

0.47

 ModelExpression

0.46

Activations Density 0.335%

references to names or entities in a historical or literary context

No Comments

No Known Activations

references to names or entities in a historical or literary context

No Comments

No Known Activations