INDEX

Explanations

hypothesized, suggested, or referenced information about the United States

oai_token-act-pair · gpt-3.5-turbo

references to the United States

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 organise

-0.70

mith

-0.70

aviour

-0.68

peat

-0.65

 whilst

-0.64

hent

-0.64

TPPStreamerBot

-0.62

 favourites

-0.62

igne

-0.61

NetMessage

-0.61

POSITIVE LOGITS

3.21

2.07

 United

1.89

US

1.68

1.47

 American

1.36

 UNITED

1.35

United

1.35

USA

1.33

Ug

1.30

Activations Density 0.057%

hypothesized, suggested, or referenced information about the United States

references to the United States

No Comments

No Known Activations