INDEX

Explanations

ethnicities or political affiliations

oai_token-act-pair · gpt-3.5-turbo

references to various ethnicities, religions, and social categories

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

Tank

-0.77

stals

-0.73

ĺħ

-0.72

Pad

-0.68

pad

-0.68

ratulations

-0.67

uders

-0.67

ricanes

-0.67

earchers

-0.66

mask

-0.66

POSITIVE LOGITS

ized

1.25

izable

1.14

ised

1.11

istic

1.11

isable

1.04

 sounding

1.04

 minded

1.01

nered

0.95

ifiable

0.94

izing

0.83

Activations Density 0.248%

ethnicities or political affiliations

references to various ethnicities, religions, and social categories

No Comments

No Known Activations