INDEX

Explanations

phrases related to hierarchy and authority

oai_token-act-pair · gpt-3.5-turbo

the conjunction "and" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Vaugh

-0.68

bub

-0.62

umbn

-0.58

æ©

-0.58

DAQ

-0.57

[*

-0.56

ä»£

-0.55

 Seym

-0.54

@@

-0.54

MRI

-0.53

POSITIVE LOGITS

rogens

0.85

 thence

0.78

 secondly

0.73

rew

0.65

rogen

0.65

 therefore

0.65

 hence

0.65

 consequently

0.64

 thus

0.63

 then

0.63

Activations Density 0.228%

phrases related to hierarchy and authority

the conjunction "and" in various contexts

No Comments

No Known Activations