INDEX

Explanations

phrases related to a particular role, position or status, especially in organizational or social contexts

oai_token-act-pair · gpt-3.5-turbo

references to various roles or statuses

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

STON

-0.72

vous

-0.68

ART

-0.67

Interstitial

-0.67

awk

-0.66

oa

-0.66

displayText

-0.65

burning

-0.62

 Flavoring

-0.62

burn

-0.62

POSITIVE LOGITS

 position

1.19

 positions

1.07

position

0.97

ngth

0.92

Position

0.89

 Position

0.83

holders

0.82

inctions

0.82

uation

0.78

osition

0.77

Activations Density 0.019%

phrases related to a particular role, position or status, especially in organizational or social contexts

references to various roles or statuses

No Comments

No Known Activations