INDEX

Explanations

references to avatars in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ی

-0.53

iastes

-0.50

esgue

-0.49

__':

-0.49

enough

-0.47

первых

-0.47

NotNull

-0.46

躇

-0.46

zelfde

-0.45

-0.43

POSITIVE LOGITS

 Avatar

1.20

Avatar

1.13

avatar

1.09

 avatar

1.05

vatar

1.03

 avatars

0.88

avatars

0.81

ATAR

0.73

berday

0.57

头像

0.50

Activations Density 0.003%

references to avatars in various contexts

No Comments

No Known Activations