INDEX

Explanations

mentions of physical injuries or risk of harm

oai_token-act-pair · gpt-3.5-turbo

references to injuries 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 @ 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

zees

-0.84

zee

-0.83

tz

-0.78

zsche

-0.77

arten

-0.74

estone

-0.68

ogle

-0.68

soDeliveryDate

-0.67

isse

-0.67

ramid

-0.65

POSITIVE LOGITS

 inflicted

1.14

 incurred

1.00

 sustained

0.91

 inflic

0.85

 suffered

0.84

 injuries

0.84

 Survivors

0.80

letal

0.78

 injury

0.76

 Injury

0.73

Activations Density 0.029%

mentions of physical injuries or risk of harm

references to injuries in various contexts

No Comments

No Known Activations