INDEX

Explanations

locations or places indicated in a text

oai_token-act-pair · gpt-3.5-turbo

locations or settings described in the text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hemy

-0.80

catentry

-0.74

henko

-0.73

cffffcc

-0.72

CLASSIFIED

-0.70

PIN

-0.66

 VIDEOS

-0.65

username

-0.64

 inval

-0.63

MSN

-0.62

POSITIVE LOGITS

 conjunction

1.19

 lieu

1.16

 accordance

1.05

patient

1.02

 vitro

0.99

efficiency

0.97

 Switzerland

0.93

 Somerset

0.93

 Pennsylvania

0.92

 southwest

0.92

Activations Density 0.348%

locations or places indicated in a text

locations or settings described in the text

No Comments

No Known Activations