INDEX

Explanations

phrases related to videos or visual recordings

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Caratter

-0.63

Autore

-0.61

 Longueur

-0.61

Flashcards

-0.60

 Outils

-0.60

Sì

-0.59

 viciss

-0.59

 Hauteur

-0.59

 charité

-0.58

Consigli

-0.57

POSITIVE LOGITS

 piemē

0.68

mī

0.62

 nepiecieš

0.60

 dårlig

0.60

ījum

0.59

islation

0.58

ipment

0.58

 footage

0.58

 mør

0.57

 izvē

0.57

Activations Density 0.271%

phrases related to videos or visual recordings

No Comments

No Known Activations