INDEX

Explanations

words indicating positive aspirations or difficult experiences

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

IVEREF

-0.98



-0.89

Geplaatst

-0.85

AutoresizingMask

-0.82

expandindo

-0.81

 AttributeSet

-0.79

BufferException

-0.78

 Roskov

-0.77

URLException

-0.76

Tikang

-0.75

POSITIVE LOGITS

to

0.83

for

0.58

 that

0.57

how

0.48

 having

0.46

 knowing

0.44

 getting

0.44

 rasanya

0.41

 bagi

0.41

ležité

0.40

Activations Density 0.197%

words indicating positive aspirations or difficult experiences

No Comments

No Known Activations