INDEX

Explanations

instances of the word "did" and its variations in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.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

abestanden

-0.54

 '\\;'

-0.47

RenderAtEndOf

-0.43

SharedCtor

-0.43

DebuggerStep

-0.42

AsUp

-0.38

PerformLayout

-0.37

ⓧ

-0.37

)++;

-0.37

__(/*!

-0.37

POSITIVE LOGITS

not

3.56

not

2.55

NOT

2.23

Not

2.17

Not

2.14

 nicht

2.00

NOT

1.93

 bukan

1.80

 не

1.77

ไม่

1.74

Activations Density 5.596%

instances of the word "did" and its variations in different contexts

No Comments

No Known Activations