INDEX

Explanations

references to law enforcement-related terms, particularly SWAT teams

oai_token-act-pair · gpt-3.5-turbo

mentions of SWAT teams and related military law enforcement activities

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

cart

-0.77

é»Ĵ

-0.70

ãĥĩãĤ£

-0.69

ezvous

-0.67

 Mits

-0.67

icative

-0.66

 Franken

-0.65

usions

-0.65

lected

-0.63

iments

-0.63

POSITIVE LOGITS

 SWAT

0.94

ILLE

0.83

antha

0.83

ilitary

0.79

ormal

0.76

auld

0.75

intendent

0.75

ELY

0.74

 Officers

0.72

ilk

0.71

Activations Density 0.016%

references to law enforcement-related terms, particularly SWAT teams

mentions of SWAT teams and related military law enforcement activities

No Comments

No Known Activations