INDEX

Explanations

negation statements about personal opinions

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

visual descriptions and atmosphere

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

 ब्रेकडाउन

-0.60

dafx

-0.57

 dieux

-0.55

ंदीखरीदारी

-0.53

 ویکی‌پدی

-0.52

 comigo

-0.50

ɝ

-0.49

Vege

-0.48

 trône

-0.48

########.

-0.47

POSITIVE LOGITS

 blush

0.35

setCellStyle

0.33

 sense

0.33

 glistening

0.32

ोर्ट

0.32

IsMutable

0.32

 uneasy

0.31

 Pink

0.31

 scorching

0.31

 Southern

0.31

Activations Density 0.025%

negation statements about personal opinions

visual descriptions and atmosphere

No Comments

No Known Activations