INDEX

Explanations

mentions of fugitives or individuals on the run from authorities

oai_token-act-pair · gpt-3.5-turbo

terms related to fugitives and criminal activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ONES

-0.85

iths

-0.83

imar

-0.71

ä»£

-0.70

eries

-0.70

renches

-0.70

ores

-0.68

andel

-0.66

eus

-0.65

uph

-0.65

POSITIVE LOGITS

fug

1.16

 fugitive

1.11

itives

1.05

querque

1.00

 felon

0.98

Fug

0.88

adoes

0.69

 felony

0.68

acious

0.68

edient

0.67

Activations Density 0.013%

mentions of fugitives or individuals on the run from authorities

terms related to fugitives and criminal activity

No Comments

No Known Activations