INDEX

Explanations

numerical representations, often pertaining to statistical data or measurements

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 betweenstory

-0.56

ACHUSET

-0.52

 resourceCulture

-0.48

 '\\;'

-0.47

 Audiodateien

-0.44

 useRouter

-0.44

arashtra

-0.43

AsUp

-0.43

AutoScaleMode

-0.42

fillType

-0.42

POSITIVE LOGITS

Autoritní

0.44

principalColumn

0.42

Билгалдахарш

0.40

 contextLoads

0.40

\{\\

0.39

 szint

0.36

RegistryLite

0.36

 nakalista

0.35

этому

0.35

锈钢

0.34

Activations Density 0.654%

numerical representations, often pertaining to statistical data or measurements

No Comments

No Known Activations