INDEX

Explanations

code snippets related to variable manipulation and function definitions

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

 noDo

-0.69

 Réponses

-0.63

tagHelperRunner

-0.63

 Lightboxes

-0.56

 estekak

-0.56

thâu

-0.56

EndContext

-0.55

 يتيمه

-0.54

تقاوى

-0.54

 Taktlose

-0.53

POSITIVE LOGITS

 remove

0.41

 removed

0.41

remove

0.40

 Remove

0.39

 adopted

0.38

Remove

0.38

atal

0.38

 inwest

0.37

 split

0.37

 Brod

0.36

Activations Density 0.018%

code snippets related to variable manipulation and function definitions

No Comments

No Known Activations