INDEX

Explanations

phrases expressing skepticism or concern regarding decisions and actions

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

 kolek

-0.34

careous

-0.33

 panoramique

-0.32

 compartilhar

-0.32

씩

-0.32

 paisagem

-0.31

 colecciones

-0.30

 ouvriers

-0.30

 commerciales

-0.29

 CanadaChoose

-0.28

POSITIVE LOGITS

 للمعارف

0.63

modb

0.59

RegistryLite

0.56

 utafitiHapana

0.56

TagMode

0.55

 estekak

0.54

copg

0.52

 AssemblyCulture

0.52

----</

0.51

ender

0.51

Activations Density 0.069%

phrases expressing skepticism or concern regarding decisions and actions

No Comments

No Known Activations