INDEX

Explanations

references to antagonists or "villains" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to villains in narratives or film

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

sterdam

-0.80

galitarian

-0.73

independent

-0.73

chedel

-0.71

independence

-0.68

DCS

-0.67

press

-0.67

porting

-0.67

glas

-0.66

oday

-0.65

POSITIVE LOGITS

ous

1.32

ously

1.10

 villain

0.98

 Bane

0.97

 villains

0.90

ess

0.86

esses

0.86

 mastermind

0.86

 Dracula

0.81

OUS

0.79

Activations Density 0.038%

references to antagonists or "villains" in various contexts

references to villains in narratives or film

No Comments

No Known Activations