INDEX

Explanations

mentions and descriptions of national matters or entities

oai_token-act-pair · gpt-3.5-turbo

references to national identity or national issues

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

Negative Logits

herty

-0.88

xual

-0.85

enhagen

-0.74

essen

-0.73

ullivan

-0.71

omething

-0.71

holes

-0.70

eva

-0.70

 Coffin

-0.70

YE

-0.69

POSITIVE LOGITS

ised

1.01

ities

0.97

istic

0.96

 Geographic

0.94

isms

0.94

 sovere

0.93

 anthem

0.92

ized

0.89

isations

0.87

 unity

0.86

Activations Density 0.037%

mentions and descriptions of national matters or entities

references to national identity or national issues

No Known Activations