INDEX

Explanations

Chinese characters or phrases, specifically those indicating a note or message for the user

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

Text after question mark or code

np_acts-logits-general · gemini-2.0-flash

names and positions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 ***!

-0.80

اریخ

-0.74

fromnode

-0.72

TintMode

-0.70

CPtr

-0.69

 iprot

-0.68

ronpa

-0.68

parsedMessage

-0.67

PyExc

-0.65

 initComponents

-0.64

POSITIVE LOGITS

 such

0.56

ISupport

0.46

anos

0.45

such

0.45

esgue

0.43

 ImGui

0.42

cope

0.42

{#

0.42

addEventListener

0.41

 enough

0.41

Activations Density 0.038%

Chinese characters or phrases, specifically those indicating a note or message for the user

Text after question mark or code

names and positions

No Comments

No Known Activations