INDEX

Explanations

names, particularly those associated with people and researchers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.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

 للاسماء

-0.84

 betweenstory

-0.68



-0.65

 '\\;'

-0.63

 Италијани

-0.60

PreferredItem

-0.59

 surla

-0.59

tagHelperRunner

-0.59

 שוליים

-0.58

:✨

-0.57

POSITIVE LOGITS

sson

0.58

son

0.54

son

0.46

ssen

0.44

sen

0.42

SON

0.40

ussen

0.40

SSON

0.38

inson

0.38

sohn

0.37

Activations Density 0.333%

names, particularly those associated with people and researchers

No Comments

No Known Activations