INDEX

Explanations

quantities and groupings of items or elements

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 âmes

-0.46

Demografia

-0.46

aktur

-0.45

ContentAsync

-0.45

 jūs

-0.43

 écrits

-0.42

 Tatsache

-0.41

대로

-0.40

 اهمیت

-0.40

 esprits

-0.39

POSITIVE LOGITS

two

0.72

 three

0.65

 different

0.63

TWO

0.60

 THREE

0.59

 dwie

0.58

 multiple

0.57

 separate

0.56

three

0.56

 dozen

0.54

Activations Density 0.566%

quantities and groupings of items or elements

No Comments

No Known Activations