INDEX

Explanations

phrases related to cyber threats and hacking techniques

oai_token-act-pair · gpt-3.5-turbo

terms related to phishing and security vulnerabilities

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

ãĥ´

-0.73

PAX

-0.67

inker

-0.65

-0.63

NCT

-0.63

semb

-0.63

semble

-0.63

 Differences

-0.63

ffen

-0.62

EVA

-0.60

POSITIVE LOGITS

WARE

0.85

 Royale

0.83

gery

0.77

 scams

0.76

 prey

0.76

ammers

0.73

ulnerable

0.73

mong

0.71

hij

0.71

oit

0.71

Activations Density 0.024%

phrases related to cyber threats and hacking techniques

terms related to phishing and security vulnerabilities

No Comments

No Known Activations