INDEX

Explanations

characters or symbols often associated with formatting or structural elements in text

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

Characters with surrounding dashes

np_acts-logits-general · gemini-2.0-flash

`@`, `‑`

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

 MainAxisSize

-0.86

 الحره

-0.86

 Infórmanos

-0.80

 ब्रेकडाउन

-0.78

 informée

-0.75

yarnpkg

-0.74

makeConstraints

-0.74

 nahilalakip

-0.73

principalColumn

-0.73

 ویکی‌پدی

-0.73

POSITIVE LOGITS

0.68

0.47

0.45

0.38

DeviceInfo

0.37

topLeft

0.36

 Boden

0.35

 bezpeč

0.34

 tóxic

0.34

 Beyer

0.33

Activations Density 0.043%

characters or symbols often associated with formatting or structural elements in text

Characters with surrounding dashes

`@`, `‑`

No Comments

No Known Activations