INDEX

Explanations

locations described by proximity in terms of distance or direction

oai_token-act-pair · gpt-3.5-turbo

references to distances and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

olin

-0.71

erity

-0.70

ole

-0.68

acus

-0.68

orian

-0.66

odic

-0.65

illions

-0.65

ospace

-0.63

 ital

-0.63

die

-0.63

POSITIVE LOGITS

 southwest

1.19

 southeast

1.17

 northeast

1.12

 northwest

1.11

 east

1.03

 west

1.02

 south

0.99

outheast

0.99

 north

0.98

 inland

0.87

Activations Density 0.144%

locations described by proximity in terms of distance or direction

references to distances and locations

No Comments

No Known Activations