INDEX

Explanations

words related to geographical locations

oai_token-act-pair · gpt-3.5-turbo

references to locations or geographical place names

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

Anonymous

-0.81

*/(

-0.81

ongyang

-0.80

vana

-0.74

iasco

-0.73

ocument

-0.72

ccording

-0.72

MpServer

-0.71

 srfAttach

-0.71

channelAvailability

-0.69

POSITIVE LOGITS

eret

0.79

shire

0.71

bury

0.70

 Abbey

0.69

furt

0.68

 Zeit

0.68

enment

0.67

aways

0.65

erer

0.63

 Christensen

0.63

Activations Density 0.104%

words related to geographical locations

references to locations or geographical place names

No Comments

No Known Activations