INDEX

Explanations

closing statements or endings

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Close" in various contexts

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

NZ

-0.76

ãĤ¼

-0.70

å¤©

-0.67

ICAN

-0.66

ä»£

-0.66

Ïī

-0.65

ULE

-0.65

åŃ

-0.62

oled

-0.61

umen

-0.61

POSITIVE LOGITS

 Close

0.89

Close

0.87

 proximity

0.84

 minded

0.83

 confinement

0.78

quarters

0.74

close

0.73

 shave

0.71

sighted

0.71

quartered

0.71

Activations Density 0.017%

closing statements or endings

instances of the word "Close" in various contexts

No Comments

No Known Activations