INDEX

Explanations

words related to cyber activities like hacking

oai_token-act-pair · gpt-3.5-turbo

terms related to hacking and cybersecurity activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Debor

-0.75

psons

-0.73

eda

-0.71

 Niet

-0.69

 livest

-0.68

ieves

-0.67

Mub

-0.66

ogue

-0.65

govtrack

-0.65

stros

-0.65

POSITIVE LOGITS

intosh

0.91

acked

0.86

ACK

0.82

heimer

0.80

luster

0.74

enzie

0.74

berry

0.69

ademic

0.69

acks

0.66

romising

0.66

Activations Density 0.023%

words related to cyber activities like hacking

terms related to hacking and cybersecurity activities

No Comments

No Known Activations