INDEX

Explanations

mentions of death incidents or discoveries of dead bodies

oai_token-act-pair · gpt-3.5-turbo

instances and descriptions of death

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

earances

-0.70

PAT

-0.67

 inferred

-0.66

 Generation

-0.65

cknowled

-0.62

 ACTIONS

-0.62

 acknowled

-0.62

ceive

-0.61

arations

-0.60

SEN

-0.59

POSITIVE LOGITS

 objectionable

0.73

bery

0.69

loe

0.68

 incrim

0.67

CVE

0.67

stadt

0.64

agascar

0.64

 compromising

0.63

fficiency

0.63

uine

0.62

Activations Density 0.318%

mentions of death incidents or discoveries of dead bodies

instances and descriptions of death

No Comments

No Known Activations