INDEX

Explanations

phrases related to specific locations, particularly those in Canada

oai_token-act-pair · gpt-3.5-turbo

place names and geographical features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

idious

-0.83

oyer

-0.78

ively

-0.77

ifice

-0.76

itbart

-0.75

 Machina

-0.72

ijn

-0.70

ischer

-0.69

mble

-0.68

onym

-0.68

POSITIVE LOGITS

 Islands

1.45

 Heights

1.10

 Isles

1.04

 Territories

1.02

 Island

0.98

 Hills

0.98

 islands

0.98

 Islanders

0.97

 County

0.97

 Highlands

0.95

Activations Density 0.172%

phrases related to specific locations, particularly those in Canada

place names and geographical features

No Comments

No Known Activations