INDEX

Explanations

forms of the verb "to be" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

です

-0.49

able

-0.47

-0.46

rgba

-0.46

ži

-0.46

 GlobalKey

-0.45

 rgba

-0.45

labelText

-0.45

 laby

-0.45

就像

-0.45

POSITIVE LOGITS

 being

1.05

 étant

0.98

oa̍t

0.88

being

0.83

oredCriteria

0.81

 BEING

0.81

 nahilalakip

0.80

Hochspringen

0.78

 kaarangay

0.77

 <=",

0.77

Activations Density 0.142%

forms of the verb "to be" in different contexts

No Comments

No Known Activations