INDEX

Explanations

words related to the color pink

oai_token-act-pair · gpt-3.5-turbo

references to the color pink

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

SPONSORED

-0.88

REC

-0.79

HAM

-0.76

=-=-=-=-

-0.74

Iraq

-0.74

REL

-0.74

UTH

-0.73

 Dresden

-0.71

Assad

-0.71

VEL

-0.69

POSITIVE LOGITS

 pink

1.25

 blob

0.95

 ribbon

0.95

 slime

0.95

 purple

0.84

 stripe

0.83

 colour

0.81

yond

0.81

 granite

0.81

bike

0.80

Activations Density 0.005%

words related to the color pink

references to the color pink

No Comments

No Known Activations