INDEX

Explanations

locations mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

geographical locations and cities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

zers

-0.74

odic

-0.71

thood

-0.70

agine

-0.68

sbm

-0.66

iosyncr

-0.65

afer

-0.64

cules

-0.63

omorph

-0.63

onym

-0.63

POSITIVE LOGITS

WA

1.02

VA

0.98

TX

0.97

TN

0.94

MA

0.89

 Ontario

0.89

KS

0.88

CA

0.88

KY

0.88

FL

0.87

Activations Density 0.063%

locations mentioned in the text

geographical locations and cities

No Comments

No Known Activations