INDEX

Explanations

repeated uses of the verb "do" in various forms

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

"Do" followed by a verb

np_acts-logits-general · gemini-2.0-flash

do followed by action or thought

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

CloseOperation

-0.91

 termica

-0.65

 soient

-0.63

 estimés

-0.61

ysacchar

-0.61

HostException

-0.59

 redenen

-0.57

ItemBackground

-0.56

 lest

-0.55

AndEndTag

-0.55

POSITIVE LOGITS

 indeed

1.01

not

1.00

 seem

0.95

NOT

0.85

 have

0.80

 however

0.78

 appear

0.76

 take

0.74

 contain

0.73

 remain

0.73

Activations Density 0.097%

repeated uses of the verb "do" in various forms

"Do" followed by a verb

do followed by action or thought

No Comments

No Known Activations