INDEX

Explanations

phrases related to incidents resulting in death

oai_token-act-pair · gpt-3.5-turbo

events involving death or fatal outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Represent

-0.68

reci

-0.64

icons

-0.62

 Cabin

-0.57

ansk

-0.56

 Dating

-0.55

 Figures

-0.55

irin

-0.54

pport

-0.54

 oneself

-0.53

POSITIVE LOGITS

 afterward

1.02

 promptly

1.00

 later

0.99

Later

0.99

 eventually

0.98

later

0.94

 thereafter

0.94

 afterwards

0.93

 ensued

0.93

sequently

0.90

Activations Density 0.695%

phrases related to incidents resulting in death

events involving death or fatal outcomes

No Comments

No Known Activations