INDEX

Explanations

specific acronyms and political terms related to political entities or movements

oai_token-act-pair · gpt-3.5-turbo

references to political parties and their activities

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

 Yellowstone

-0.83

 Wells

-0.75

 Fleming

-0.75

Hep

-0.75

 Whitney

-0.75

 Doyle

-0.74

obyl

-0.74

 Ness

-0.74

 Hazard

-0.73

 Dickens

-0.71

POSITIVE LOGITS

 leader

1.11

Leader

1.10

 stronghold

1.03

 militant

0.99

 insurgency

0.97

 faction

0.96

 leadership

0.96

 rebels

0.95

affiliated

0.94

 guerrilla

0.94

Activations Density 0.167%

specific acronyms and political terms related to political entities or movements

references to political parties and their activities

No Comments

No Known Activations