INDEX

Explanations

references to the concept of "any" or "anything."

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

SequentialGroup

-0.66

VIAF

-0.63

RTSN

-0.54

DrawerToggle

-0.52

elemField

-0.52

Personendaten

-0.50

osoba

-0.49

adaptiveStyles

-0.48

 TextInputType

-0.48

awtextra

-0.47

POSITIVE LOGITS

No

0.57

 extra

0.56

 none

0.55

no

0.55

 additional

0.53

bergen

0.51

None

0.51

 more

0.50

no

0.50

No

0.50

Activations Density 0.027%

references to the concept of "any" or "anything."

No Comments

No Known Activations