INDEX

Explanations

numeric image captions typically containing two numerical values separated by a slash '/' character

oai_token-act-pair · gpt-3.5-turbo

references to image counts or identifiers in a context related to visual media or articles

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

worms

-0.66

won

-0.66

anism

-0.61

 acupuncture

-0.58

ually

-0.58

went

-0.56

uality

-0.55

folk

-0.55

 oper

-0.54

nia

-0.54

POSITIVE LOGITS

 Caption

0.74

 Thumbnails

0.65

 Paramount

0.62

Pinterest

0.62

urden

0.61

vertisements

0.61

ourke

0.60

ECTION

0.60

arger

0.60

Loading

0.60

Activations Density 0.031%

numeric image captions typically containing two numerical values separated by a slash '/' character

references to image counts or identifiers in a context related to visual media or articles

No Comments

No Known Activations