INDEX

Explanations

words indicating conclusion or termination

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

EnableWeb

-0.62

 daarvan

-0.59

 ransom

-0.58

MLLoader

-0.58

retweeted

-0.57

iented

-0.56

braio

-0.55

よいよ

-0.55

<bos>

-0.54

لينكات

-0.53

POSITIVE LOGITS

 ended

1.20

 ends

1.08

 ending

1.06

 Ended

0.93

结束

0.92

 Ending

0.92

Ended

0.91

 terminated

0.91

 Ends

0.89

 stopped

0.83

Activations Density 0.144%

words indicating conclusion or termination

No Comments

No Known Activations