INDEX

Explanations

phrases related to the opening or establishment of businesses or locations

oai_token-act-pair · gpt-3.5-turbo

locations and mentions of places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

TPP

-0.72

death

-0.70

Syrian

-0.70

abuse

-0.68

AFP

-0.68

azeera

-0.67

hedon

-0.67

immune

-0.65

pol

-0.65

Android

-0.64

POSITIVE LOGITS

 downtown

1.50

 Downtown

1.35

 Chinatown

1.19

 conjunction

1.12

 town

1.09

 Arlington

1.08

 lieu

1.04

 neighborhoods

0.99

 Ballard

0.99

Activations Density 0.205%

phrases related to the opening or establishment of businesses or locations

locations and mentions of places

No Comments

No Known Activations