INDEX

Explanations

instances of the word "who."

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

"who" followed by a verb

np_acts-logits-general · gemini-2.0-flash

who follows

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ArrowToggle

-0.66

RenderAtEndOf

-0.62

 snowing

-0.59

 betweenstory

-0.58

 utafitiHapana

-0.57

 circulation

-0.55

SceneManagement

-0.53

 gyerek

-0.53

خانه

-0.52

 defaultstate

-0.52

POSITIVE LOGITS

oping

0.87

who

0.84

 else

0.72

 whom

0.72

Who

0.70

has

0.67

WHO

0.66

Who

0.65

Whoosh

0.63

ltk

0.60

Activations Density 0.078%

instances of the word "who."

"who" followed by a verb

who follows

No Comments

No Known Activations