INDEX

Explanations

names of places or locations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "data."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

enegger

-0.72

paren

-0.70

tails

-0.68

taining

-0.64

 convict

-0.64

sted

-0.63

Ö¼

-0.61

false

-0.59

starter

-0.59

 premise

-0.59

POSITIVE LOGITS

iba

0.96

eus

0.95

Ã±a

0.94

ata

0.92

hedral

0.91

hea

0.88

illac

0.87

isy

0.87

Ã±

0.86

ichi

0.83

Activations Density 0.009%

names of places or locations

occurrences of the word "data."

No Comments

No Known Activations