INDEX

Explanations

threatening or harmful situations that need to be averted

oai_token-act-pair · gpt-3.5-turbo

phrases related to prevention

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ammy

-0.84

eah

-0.72

edded

-0.69

ebus

-0.64

aturday

-0.62

swick

-0.62

 Featured

-0.62

enegger

-0.62

geist

-0.60

 Style

-0.60

POSITIVE LOGITS

ative

0.98

 detection

0.86

 inhib

0.81

ively

0.76

 pregnancies

0.73

 regress

0.73

ministic

0.73

 duplicate

0.72

00200000

0.68

 obstruct

0.67

Activations Density 0.033%

threatening or harmful situations that need to be averted

phrases related to prevention

No Comments

No Known Activations