INDEX

Explanations

negations and indicators of falsehood or non-fulfillment

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

False followed by numbers or symbols

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

WebElementEntity

-0.65

 betweenstory

-0.56

 تضيفلها

-0.52

MemoryWarning

-0.52

AutoScaleMode

-0.51

TestingModule

-0.51

 '\\;'

-0.50

 ModelExpression

-0.49

verwijspagina

-0.47

PreferredItem

-0.47

POSITIVE LOGITS

tab

0.34

ork

0.33

tab

0.33

事

0.32

kpi

0.32

 imposible

0.32

rungsseite

0.31

 pure

0.31

Amt

0.30

stan

0.30

Activations Density 0.239%

negations and indicators of falsehood or non-fulfillment

False followed by numbers or symbols

No Comments

No Known Activations