INDEX

Explanations

words related to violent incidents resulting in death

oai_token-act-pair · gpt-3.5-turbo

statistics related to casualties from violent events or attacks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Printed

-0.88

soDeliveryDate

-0.86

ellar

-0.79

SPONSORED

-0.77

 disclaimer

-0.76

WARN

-0.76

soType

-0.74

XM

-0.70

oyer

-0.70

published

-0.70

POSITIVE LOGITS

 unarmed

1.02

 senseless

0.90

 innocent

0.85

 rampage

0.83

 injuring

0.82

 arson

0.82

 landslide

0.82

 hostages

0.81

 fleeing

0.80

 violently

0.80

Activations Density 0.401%

words related to violent incidents resulting in death

statistics related to casualties from violent events or attacks

No Comments

No Known Activations