INDEX

Explanations

references to people visiting different places or individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "visited" or related forms indicating past visits or interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

xual

-0.84

dp

-0.78

frac

-0.78

claimer

-0.73

gex

-0.73

Ĳ

-0.70

rama

-0.69

direction

-0.68

reg

-0.67

ubi

-0.67

POSITIVE LOGITS

 museums

0.97

 Hiroshima

0.92

 Disneyland

0.91

 Cuba

0.85

 Arlington

0.85

 abroad

0.83

 temples

0.82

 relatives

0.82

 Taiwan

0.82

 Chinatown

0.81

Activations Density 0.062%

references to people visiting different places or individuals

instances of the word "visited" or related forms indicating past visits or interactions

No Comments

No Known Activations