INDEX

Explanations

phrases related to actions or events happening or occurring

oai_token-act-pair · gpt-3.5-turbo

references to upcoming events or releases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

catentry

-0.56

Vers

-0.49

[|

-0.46

 relates

-0.45

 Shoot

-0.45

 Practices

-0.43

 freezes

-0.43

doms

-0.43

scenes

-0.43

 Culture

-0.42

POSITIVE LOGITS

enced

0.58

lycer

0.57

cedented

0.53

roo

0.51

emonium

0.51

roud

0.49

ailable

0.48

kaya

0.48

glers

0.47

.–

0.47

Activations Density 7.158%

phrases related to actions or events happening or occurring

references to upcoming events or releases

No Comments

No Known Activations