INDEX

Explanations

words related to specific proper nouns, particularly related to historical, medical, or governmental contexts

oai_token-act-pair · gpt-3.5-turbo

references to Rhodes and Uranium

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 ILCS

-0.74

bernatorial

-0.69

ASE

-0.68

à¨

-0.68

 Archangel

-0.66

aneous

-0.64

à©

-0.64

ISM

-0.62

 Islands

-0.62

ONT

-0.62

POSITIVE LOGITS

kaya

0.77

hern

0.77

peria

0.76

umen

0.76

ko

0.75

iop

0.75

iation

0.75

essee

0.74

kh

0.73

kees

0.72

Activations Density 0.079%

words related to specific proper nouns, particularly related to historical, medical, or governmental contexts

references to Rhodes and Uranium

No Comments

No Known Activations