INDEX

Explanations

specific identifiers and names within a variety of contexts

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

 تضيفلها

-0.71

 الحره

-0.69

AndEndTag

-0.68

 للمعارف

-0.66

存于互联网档案馆

-0.66

 فريبيس

-0.63

تقاوى

-0.58

tagHelperRunner

-0.54

Lähteet

-0.53

 كومونز

-0.53

POSITIVE LOGITS

GEBURTSDATUM

0.42

uxxxx

0.39

SBATCH

0.39

tel

0.38

 findViewById

0.35

 moreover

0.35

 incidentally

0.32

MeasureSpec

0.31

 another

0.31

 Bahn

0.30

Activations Density 0.575%

specific identifiers and names within a variety of contexts

No Comments

No Known Activations