INDEX

Explanations

repeated conjunctions or phrases indicating connection or addition

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.12.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

chanind/pile-uncopyrighted-gemma-1024-abbrv-1B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Personendaten

-0.87

dafx

-0.82

Autoritní

-0.82

]")]

-0.77

ⓧ

-0.75

 мәкал

-0.73

 snippetHide

-0.72

ьаж

-0.70

ThemeOverlay

-0.70

 Infórmanos

-0.68

POSITIVE LOGITS

 former

0.69

 Nowak

0.67

former

0.63

 meist

0.61

CreateModel

0.60

èze

0.59

 senior

0.58

 door

0.57

 multi

0.57

ving

0.56

Activations Density 0.050%

repeated conjunctions or phrases indicating connection or addition

No Comments

No Known Activations