INDEX

Explanations

extreme adjectives describing severity, complexity, or power

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

ReusableCell

-0.81

 мәкалә

-0.67

 MainAxisSize

-0.66

 NSCoder

-0.64

 كومونز

-0.62

Билгалдахарш

-0.62

الإنجليزية

-0.60

 TestBed

-0.60

 Seeder

-0.59

colhead

-0.59

POSITIVE LOGITS

 înc

1.12

it

0.92

 that

0.73

 sehingga

0.72

so

0.65

you

0.62

 tellement

0.62

 tantos

0.56

 zoveel

0.56

 zelfs

0.56

Activations Density 0.112%

extreme adjectives describing severity, complexity, or power

No Comments

No Known Activations