INDEX

Explanations

words related to safety

oai_token-act-pair · gpt-3.5-turbo

references to safety 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 @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ovi

-0.75

opus

-0.69

txt

-0.67

gd

-0.66

zzo

-0.65

itus

-0.64

agne

-0.64

 Married

-0.64

ago

-0.63

Saharan

-0.62

POSITIVE LOGITS

 safety

3.94

safety

3.34

 Safety

2.93

Safety

2.92

 safer

1.77

 safe

1.69

afety

1.67

saf

1.60

SAF

1.58

 safest

1.54

Activations Density 0.022%

words related to safety

references to safety in various contexts

No Comments

No Known Activations