INDEX

Explanations

mentions of individuals in positions of authority or leadership

oai_token-act-pair · gpt-3.5-turbo

references to leadership positions or titles

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

Pwr

-0.75

urses

-0.70

 Fusion

-0.69

zl

-0.66

ogene

-0.64

usion

-0.64

berra

-0.63

 belly

-0.63

 Leilan

-0.63

nep

-0.63

POSITIVE LOGITS

boards

1.10

board

0.97

doms

0.92

leader

0.91

aroo

0.90

pins

0.90

pin

0.83

wcs

0.81

pieces

0.79

stration

0.78

Activations Density 0.031%

mentions of individuals in positions of authority or leadership

references to leadership positions or titles

No Comments

No Known Activations