INDEX

Explanations

instances of people standing or sitting in various locations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in" indicating location or context

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

catentry

-0.73

soever

-0.72

%%

-0.68

ŀ

-0.67

=""

-0.66

username

-0.65

etheless

-0.64

casters

-0.64

 lodged

-0.63

¿½

-0.62

POSITIVE LOGITS

ordinate

1.13

 lieu

1.07

 accordance

1.06

 vain

1.04

 conjunction

1.03

 unison

0.98

 tandem

0.97

 front

0.95

strument

0.94

 relation

0.92

Activations Density 0.286%

instances of people standing or sitting in various locations

occurrences of the word "in" indicating location or context

No Comments

No Known Activations