INDEX

Explanations

words and phrases associated with software or technology endorsements

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

Followed by "one" or "quite"

np_acts-logits-general · gemini-2.0-flash

perhaps one of

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ddelweddau

-0.81

MessageOf

-0.69

tonode

-0.66

 للاسماء

-0.66

parsedMessage

-0.65

}")

-0.62

 архивлан

-0.62

 Reſ

-0.61

SourceChecksum

-0.60

 betweenstory

-0.59

POSITIVE LOGITS

 arguably

1.98

 perhaps

1.86

 probably

1.76

Probably

1.67

 Probably

1.62

Perhaps

1.59

 Perhaps

1.58

perhaps

1.52

one

1.50

probably

1.49

Activations Density 0.439%

words and phrases associated with software or technology endorsements

Followed by "one" or "quite"

perhaps one of

No Comments

No Known Activations