INDEX

Explanations

instances of words related to harm, injury, or destruction

oai_token-act-pair · gpt-3.5-turbo

references to casualties and related terminology in the context of war and conflict

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

perm

-0.83

ramid

-0.74

itudinal

-0.69

por

-0.67

odes

-0.67

erson

-0.64

 enlightened

-0.63

appa

-0.62

 Collider

-0.62

angled

-0.62

POSITIVE LOGITS

 casualties

1.17

 bystanders

0.95

 casualty

0.93

 inflicted

0.87

 incurred

0.83

 suffered

0.79

 Victims

0.76

 victims

0.74

 civilians

0.73

 losses

0.73

Activations Density 0.022%

instances of words related to harm, injury, or destruction

references to casualties and related terminology in the context of war and conflict

No Comments

No Known Activations