INDEX

Explanations

references to the quality of a job being done

oai_token-act-pair · gpt-3.5-turbo

phrases that evaluate the quality of work or performance

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

fty

-0.71

irs

-0.68

NetMessage

-0.66

Phi

-0.65

iku

-0.62

 fears

-0.61

veh

-0.60

bidden

-0.60

oggles

-0.59

ories

-0.58

POSITIVE LOGITS

 ethic

0.80

VALUE

0.71

job

0.70

ufact

0.69

 captcha

0.69

galitarian

0.68

ixel

0.67

 description

0.67

 portfolio

0.66

vre

0.66

Activations Density 0.032%

references to the quality of a job being done

phrases that evaluate the quality of work or performance

No Comments

No Known Activations