INDEX

Explanations

statements or mentions of something being manipulated or unfairly influenced

oai_token-act-pair · gpt-3.5-turbo

the concept of election rigging

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

exist

-0.79

irs

-0.73

clamation

-0.72

birth

-0.71

grave

-0.71

gap

-0.70

ku

-0.68

sign

-0.68

Appearance

-0.68

Asia

-0.68

POSITIVE LOGITS

 rigged

1.77

 rigging

1.44

eering

0.97

 manipulated

0.81

rig

0.78

 rotten

0.76

 manip

0.72

 barg

0.71

 reel

0.70

 gamb

0.70

Activations Density 0.006%

statements or mentions of something being manipulated or unfairly influenced

the concept of election rigging

No Comments

No Known Activations