INDEX

Explanations

locations around the world, with a focus on specific regions and countries

oai_token-act-pair · gpt-3.5-turbo

geographic locations and 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

 bandwagon

-0.60

 Europeans

-0.58

":"/

-0.58

=/

-0.57

GHC

-0.56

 Bucc

-0.55

 Moment

-0.55

 Ukrainians

-0.54

Tee

-0.53

POSITIVE LOGITS

 suburb

0.92

 township

0.92

ruary

0.90

 Township

0.87

 province

0.86

fecture

0.85

 countryside

0.81

 city

0.80

's

0.79

 bordering

0.77

Activations Density 0.210%

locations around the world, with a focus on specific regions and countries

geographic locations and places

No Comments

No Known Activations