INDEX

Explanations

variations of the word "true."

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

 Monfieur

-0.70

 rodríguez

-0.69

 ListTile

-0.67

 purpoſe

-0.67

 LXXX

-0.66

 itſelf

-0.65

 nepiecieš

-0.65

 need

-0.62

ListTile

-0.62

 fuper

-0.62

POSITIVE LOGITS

:✨

0.88

trane

0.82

PRNewswire

0.73

MessageTagHelper

0.72

 AttributeSet

0.71

onAttach

0.71

 OMITBAD

0.69

VIAF

0.69

]")]

0.68

 '\\;'

0.68

Activations Density 0.017%

variations of the word "true."

No Comments

No Known Activations