INDEX

Explanations

phrases related to news articles, political figures, and events

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

 swarovski

-0.82

 Septembre

-0.70

Rgds

-0.69

 bahay

-0.68

 Octobre

-0.68

 Ename

-0.66

 haup

-0.63

 tanong

-0.63

 considération

-0.63

 maraming

-0.63

POSITIVE LOGITS

 Images

0.72

 Image

0.60

 images

0.59

 image

0.58

Images

0.57

images

0.55

 IMAGES

0.55

 Photos

0.55

Image

0.54

 Photographer

0.51

Activations Density 0.066%

phrases related to news articles, political figures, and events

No Comments

No Known Activations