INDEX

Explanations

references to destruction or removal of objects

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

Negative Logits

 sick

-0.57

 deta

-0.55

 withdraw

-0.52

 resigned

-0.52

ibat

-0.52

 lockout

-0.52

 newOwner

-0.52

 vacated

-0.51

arrête

-0.51

 withdrawal

-0.51

POSITIVE LOGITS

AndEndTag

0.67

 dinam

0.60

 урна

0.59

execSQL

0.56

UrlResolution

0.55

 détru

0.55

PageIndex

0.54

NonQuery

0.53

ishing

0.52

 MainAxisSize

0.52

Activations Density 0.122%

references to destruction or removal of objects

No Known Activations