INDEX

Explanations

references to designers or design-related keywords

oai_token-act-pair · gpt-3.5-turbo

references to fashion designers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Pradesh

-0.79

ibaba

-0.73

cffffcc

-0.72

lehem

-0.70

nikov

-0.67

avez

-0.65

zees

-0.65

olit

-0.64

ilitary

-0.63

 sovere

-0.63

POSITIVE LOGITS

urally

0.96

 designer

0.93

paces

0.90

yout

0.79

 designing

0.74

smanship

0.73

Artist

0.71

 jeans

0.70

 designers

0.70

pace

0.70

Activations Density 0.021%

references to designers or design-related keywords

references to fashion designers

No Comments

No Known Activations