INDEX

Explanations

proper nouns related to locations, such as states and universities

oai_token-act-pair · gpt-3.5-turbo

references to geographic locations and associated events or statistics

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

 agre

-0.80

xual

-0.75

 incent

-0.74

andowski

-0.74

atron

-0.71

wip

-0.70

ittle

-0.69

 answ

-0.68

berger

-0.68

awa

-0.68

POSITIVE LOGITS

Population

1.01

ccording

0.94

Associated

0.90

Introduced

0.83

................................

0.82

ãĥŁ

0.81

Notable

0.81

Lago

0.79

Ø§ÙĦ

0.78

Country

0.78

Activations Density 0.194%

proper nouns related to locations, such as states and universities

references to geographic locations and associated events or statistics

No Comments

No Known Activations