INDEX

Explanations

phrases indicating updates or notifications about changes and news

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

updates, path, get, when

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 AppCompatTheme

-0.49

JNIEnv

-0.48

 AttributeSet

-0.44

 HasFactory

-0.41

TagMode

-0.41

TagHelper

-0.40

########.

-0.40

MeasureSpec

-0.40

 representation

-0.39

MessageOf

-0.39

POSITIVE LOGITS

 ModelExpression

0.58

 werde

0.41

 dibutuhkan

0.40

ftagPool

0.39

posted

0.38

ugas

0.36

bunya

0.36

 >=",

0.36

 nauki

0.36

bitField

0.36

Activations Density 0.090%

phrases indicating updates or notifications about changes and news

updates, path, get, when

No Comments

No Known Activations