INDEX

Explanations

mentions of specific locations or regions around the world

oai_token-act-pair · gpt-3.5-turbo

references to locations and conditions in urban and rural contexts

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

DragonMagazine

-0.97

unity

-0.87

Detect

-0.80

MSN

-0.78

tle

-0.78

mem

-0.71

ãĤ¤ãĥĪ

-0.71

awn

-0.70

ãĥĭ

-0.70

cam

-0.67

POSITIVE LOGITS

 wealthier

1.36

 poorer

1.20

 Appalach

1.20

 regions

1.18

 industrialized

1.17

 affluent

1.15

 cities

1.12

 disadvantaged

1.12

effic

1.10

 developing

1.10

Activations Density 0.298%

mentions of specific locations or regions around the world

references to locations and conditions in urban and rural contexts

No Comments

No Known Activations