INDEX

Explanations

references to clarity and usefulness in a context

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

 '\\;'

-0.64

RegistryLite

-0.64

sizeCache

-0.60

EndInit

-0.56

kháu

-0.56

 TextAppearance

-0.51

ByPrimaryKey

-0.51

:✨

-0.51

 Anſ

-0.51

UnknownFields

-0.50

POSITIVE LOGITS

 surla

0.34

keln

0.34

eradish

0.32

절

0.32

 olma

0.31

 lets

0.30

 excuse

0.30

تقاوى

0.29

 good

0.29

 pitch

0.29

Activations Density 0.107%

references to clarity and usefulness in a context

No Comments

No Known Activations