INDEX

Explanations

instructions or recommendations to perform specific actions

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize reminders or suggestions

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

gery

-0.76

alist

-0.74

impl

-0.72

rock

-0.69

pher

-0.68

cience

-0.66

folk

-0.66

alt

-0.64

heres

-0.62

bern

-0.62

POSITIVE LOGITS

 Availability

0.71

icio

0.71

 beforehand

0.69

quished

0.66

 Siren

0.65

 thous

0.63

reau

0.62

 caveats

0.62

!:

0.62

yip

0.61

Activations Density 0.068%

instructions or recommendations to perform specific actions

phrases that emphasize reminders or suggestions

No Comments

No Known Activations