INDEX

Explanations

mentions of specific locations or buildings

oai_token-act-pair · gpt-3.5-turbo

phrases related to spaces and buildings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

obin

-0.67

 Niet

-0.65

 Samoa

-0.59

è¦ļéĨĴ

-0.59

aughs

-0.58

_.

-0.57

 aven

-0.56

flu

-0.55

 Rebell

-0.54

 Afgh

-0.54

POSITIVE LOGITS

Located

0.76

taboola

0.75

 Doors

0.75

 floors

0.73

 doors

0.71

 Located

0.71

 spacious

0.70

 parks

0.69

 Gardens

0.69

 Rooms

0.68

Activations Density 1.345%

mentions of specific locations or buildings

phrases related to spaces and buildings

No Comments

No Known Activations