INDEX

Explanations

places and locations with specific names

oai_token-act-pair · gpt-3.5-turbo

references to places and events in Berlin

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Cyrus

-0.80

 Mississ

-0.71

 Cherokee

-0.68

 Indian

-0.65

 Cree

-0.65

luaj

-0.64

 Indians

-0.64

 Yemeni

-0.64

 lime

-0.64

 Fairfax

-0.63

POSITIVE LOGITS

heit

1.45

mann

1.45

bach

1.44

kamp

1.43

meier

1.43

haus

1.39

hof

1.39

lich

1.38

wagen

1.38

furt

1.37

Activations Density 0.282%

places and locations with specific names

references to places and events in Berlin

No Comments

No Known Activations