INDEX

Explanations

references to United States government and policy actions

oai_token-act-pair · gpt-3.5-turbo

references to the United States and its various policies or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 organisers

-0.64

 EDIT

-0.63

nir

-0.61

 rumours

-0.60

 Comet

-0.60

 STATS

-0.60

 Nanto

-0.59

 Mermaid

-0.59

 Carbuncle

-0.59

 Metatron

-0.59

POSITIVE LOGITS

Â¢

0.84

insured

0.77

taboola

0.77

 America

0.76

arijuana

0.73

amacare

0.72

America

0.72

CLASS

0.71

astical

0.70

regulated

0.70

Activations Density 0.152%

references to United States government and policy actions

references to the United States and its various policies or conditions

No Comments

No Known Activations