INDEX

Explanations

people's names and specific roles or positions they hold

oai_token-act-pair · gpt-3.5-turbo

mentions of individuals connected to legal or political contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 cautiously

-0.59

 strategically

-0.59

 carefully

-0.58

 archaeologists

-0.58

 logically

-0.56

etheless

-0.56

)=(

-0.56

 astronomers

-0.55

 archived

-0.55

 spoilers

-0.54

POSITIVE LOGITS

$.

0.76

)).

0.62

ilda

0.62

 Olympia

0.57

zar

0.56

ãĤ©

0.55

 Avenue

0.55

)),

0.54

inton

0.54

eva

0.54

Activations Density 1.111%

people's names and specific roles or positions they hold

mentions of individuals connected to legal or political contexts

No Comments

No Known Activations