INDEX

Explanations

mentions of physical injuries

oai_token-act-pair · gpt-3.5-turbo

references to injuries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

arten

-0.76

tz

-0.75

soDeliveryDate

-0.73

zee

-0.71

ullivan

-0.71

ramid

-0.67

arity

-0.67

estone

-0.67

az

-0.66

adian

-0.66

POSITIVE LOGITS

 inflicted

1.28

 incurred

1.10

 sustained

1.06

 suffered

1.01

letal

0.90

 injuries

0.89

 inflic

0.84

 caused

0.81

 wounds

0.78

 victims

0.78

Activations Density 0.042%

mentions of physical injuries

references to injuries

No Comments

No Known Activations