INDEX

Explanations

phrases that indicate disclosure or references to previous work and methodologies

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

Followed by an example

np_acts-logits-general · gemini-2.0-flash

for example

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

sizeCache

-0.56

adaptiveStyles

-0.48

 Gros

-0.47

ftagPool

-0.45

السكان

-0.44

StructEnd

-0.44

ंदीखरीदारी

-0.43

 chi̍t

-0.42

 Italijani

-0.41

留

-0.40

POSITIVE LOGITS

 например

1.36

 misalnya

1.21

 beispielsweise

1.16

 například

1.16

 bijvoorbeeld

1.13

例えば

1.09

 např

1.09

 például

1.04

Например

1.04

たとえば

1.04

Activations Density 0.691%

phrases that indicate disclosure or references to previous work and methodologies

Followed by an example

for example

No Comments

No Known Activations