INDEX

Explanations

references to law enforcement or police actions

oai_token-act-pair · gpt-3.5-turbo

mentions of police

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

reciation

-0.75

pring

-0.72

(<

-0.71

inav

-0.70

 palate

-0.69

tle

-0.69

peak

-0.69

folio

-0.68

aments

-0.67

hesis

-0.67

POSITIVE LOGITS

 police

3.40

police

3.07

 Police

2.83

Police

2.70

 cops

2.61

 policemen

2.42

 policeman

2.33

 policing

2.01

 RCMP

1.99

 NYPD

1.90

Activations Density 0.057%

references to law enforcement or police actions

mentions of police

No Comments

No Known Activations