INDEX

Explanations

terms related to hacking or cyber attacks

oai_token-act-pair · gpt-3.5-turbo

references to hacking activities and related terminology

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

oran

-0.72

oplan

-0.72

âĢ¢âĢ¢âĢ¢âĢ¢

-0.65

 Towns

-0.62

obe

-0.60

orem

-0.57

Pres

-0.57

ŃĶ

-0.57

anguage

-0.57

 FORM

-0.57

POSITIVE LOGITS

intosh

1.28

athon

0.94

aday

0.93

paces

0.87

RF

0.85

romising

0.83

ney

0.80

attack

0.80

ileaks

0.77

netic

0.75

Activations Density 0.060%

terms related to hacking or cyber attacks

references to hacking activities and related terminology

No Comments

No Known Activations