INDEX

Explanations

interrogative phrases, particularly those asking questions about understanding or seeking clarification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 persino

-0.51

 ALSO

-0.50

TOO

-0.48

Even

-0.48

 també

-0.47

ALWAYS

-0.47

だけでなく

-0.46

はもちろん

-0.45

Too

-0.45

 chiunque

-0.45

POSITIVE LOGITS

oredCriteria

0.73

 <=",

0.72

 exactly

0.70

లాలు

0.68

exactly

0.67

 насељу

0.66

?',

0.65

hyrchwyd

0.63

 exac

0.63

 happened

0.63

Activations Density 0.147%

interrogative phrases, particularly those asking questions about understanding or seeking clarification

No Comments

No Known Activations