INDEX

Explanations

numbers within text

oai_token-act-pair · gpt-3.5-turbo

numerical or chronological references in a text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

heit

-0.70

boro

-0.64

 diseng

-0.63

ardless

-0.62

ocial

-0.60

umni

-0.59

abouts

-0.59

isexual

-0.56

activity

-0.54

assador

-0.53

POSITIVE LOGITS

Scroll

0.75

However

0.69

 Debor

0.68

reditary

0.68

ECK

0.66

Until

0.65

ÂłÂłÂł

0.63

Lear

0.63

Specifically

0.63

SEE

0.63

Activations Density 0.627%

numbers within text

numerical or chronological references in a text

No Comments

No Known Activations