INDEX

Explanations

references to different professions and nationalities

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

+#+#

-0.69

piram

-0.54

RegressionTest

-0.53

AFX

-0.50

دانشنامهٔ

-0.50

IMPORTED

-0.50

Sach

-0.49

 slams

-0.47

PostInfinity

-0.46

 Bulletin

-0.46

POSITIVE LOGITS

 habet

0.46

balleur

0.41

ngdoc

0.40

 profesión

0.39

 avoient

0.39

 Waray

0.39

 écrivain

0.37

 laiko

0.37

Parcelize

0.35

 chrétiens

0.34

Activations Density 0.025%

references to different professions and nationalities

No Comments

No Known Activations