INDEX

Explanations

issues of crime and security

oai_token-act-pair · gpt-3.5-turbo

references to illegal activities or criminal investigations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

izons

-0.87

tarians

-0.79

iences

-0.79

 evaluates

-0.73

isans

-0.73

aez

-0.73

ependence

-0.72

iors

-0.72

stars

-0.71

 spectators

-0.71

POSITIVE LOGITS

 scam

1.31

 hoax

1.22

 ransomware

1.21

 fraudulent

1.20

 malware

1.19

 bogus

1.16

 phony

1.13

 scheme

1.11

 fake

1.09

 leak

1.07

Activations Density 0.390%

issues of crime and security

references to illegal activities or criminal investigations

No Comments

No Known Activations