INDEX

Explanations

phrases related to followers or worshipers of a specific person or ideology

oai_token-act-pair · gpt-3.5-turbo

references to social media followers

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

ces

-0.93

ced

-0.74

 circumstance

-0.73

cer

-0.73

 Genocide

-0.68

OUT

-0.68

pmwiki

-0.66

posing

-0.65

ospital

-0.62

 Nurse

-0.62

POSITIVE LOGITS

 followers

1.35

hip

1.21

 follower

1.09

ollower

0.99

 adherents

0.92

 adherent

0.84

 disciples

0.80

hips

0.78

lia

0.77

isSpecialOrderable

0.74

Activations Density 0.010%

phrases related to followers or worshipers of a specific person or ideology

references to social media followers

No Comments

No Known Activations