INDEX

Explanations

references to political power and leadership, particularly in the context of a presidency

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "presidency."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Ku

-0.75

Thom

-0.72

srf

-0.69

bidden

-0.69

Quality

-0.69

izen

-0.69

 IMAGES

-0.69

Ô

-0.68

âĵĺ

-0.67

ļéĨĴ

-0.67

POSITIVE LOGITS

 presidency

1.11

 Presidency

1.06

manship

0.86

OTUS

0.84

 office

0.78

velt

0.73

 writ

0.72

 Office

0.72

 suite

0.71

matically

0.71

Activations Density 0.019%

references to political power and leadership, particularly in the context of a presidency

occurrences of the word "presidency."

No Comments

No Known Activations