INDEX

Explanations

phrases related to physical spaces like "locker room."

oai_token-act-pair · gpt-3.5-turbo

references to locker rooms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

alez

-0.82

reme

-0.69

akuya

-0.67

ammed

-0.64

nom

-0.63

eff

-0.63

cale

-0.62

rendered

-0.62

DonaldTrump

-0.61

eva

-0.60

POSITIVE LOGITS

 locker

1.25

 rooms

0.85

 room

0.84

room

0.83

lain

0.76

 clubhouse

0.76

 drawer

0.73

bie

0.73

 rehearsal

0.72

staff

0.72

Activations Density 0.008%

phrases related to physical spaces like "locker room."

references to locker rooms

No Comments

No Known Activations