INDEX

Explanations

numerical expressions indicating a specific portion or fraction of a group

oai_token-act-pair · gpt-3.5-turbo

phrases indicating statistical occurrences or proportions

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

-0.86

adm

-0.63

 disple

-0.62

LOG

-0.61

 anch

-0.61

ben

-0.61

channelAvailability

-0.60

Wiki

-0.59

vag

-0.59

 captcha

-0.58

POSITIVE LOGITS

patient

1.18

organic

1.13

clusively

1.06

ordinate

1.03

clusions

0.99

humane

0.96

animate

0.96

roads

0.95

effic

0.94

ggles

0.94

Activations Density 0.106%

numerical expressions indicating a specific portion or fraction of a group

phrases indicating statistical occurrences or proportions

No Comments

No Known Activations