INDEX

Explanations

verbs prefixed with 're-' indicating some form of action being repeated or done again

oai_token-act-pair · gpt-3.5-turbo

words related to starting, reviewing, or changing processes

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

elo

-0.68

aghan

-0.63

Bir

-0.58

aley

-0.57

 Lily

-0.57

erity

-0.57

TW

-0.56

 unsur

-0.56

gram

-0.55

Writing

-0.54

POSITIVE LOGITS

issance

1.23

xual

0.95

itory

0.94

owship

0.90

ction

0.90

burse

0.85

ctions

0.80

foundland

0.79

irmation

0.77

enment

0.76

Activations Density 0.057%

verbs prefixed with 're-' indicating some form of action being repeated or done again

words related to starting, reviewing, or changing processes

No Comments

No Known Activations