INDEX

Explanations

references to National Geographic

oai_token-act-pair · gpt-3.5-turbo

references to National Geographic and its associated content

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

ardless

-0.84

ammy

-0.80

unts

-0.78

raising

-0.76

ding

-0.75

ework

-0.74

bub

-0.74

ded

-0.73

ione

-0.72

rons

-0.72

POSITIVE LOGITS

 Channel

0.98

 Adventures

0.79

 Geographic

0.79

 Maps

0.76

 Explorer

0.75

Maps

0.74

 Warfare

0.73

 Editors

0.73

arium

0.73

Map

0.71

Activations Density 0.006%

references to National Geographic

references to National Geographic and its associated content

No Comments

No Known Activations