INDEX

Explanations

phrases indicating precision or spatial relationships such as "to the north" or "about 30 kilometers"

oai_token-act-pair · gpt-3.5-turbo

prepositions and phrases indicating direction or position

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

interrupted

-0.53

NESS

-0.48

owan

-0.47

 selves

-0.45

orer

-0.44

/,

-0.44

/+

-0.44

arial

-0.42

oa

-0.42

ials

-0.41

POSITIVE LOGITS

the

1.14

the

1.09

THE

0.81

The

0.70

tha

0.68

é¾

0.64

THE

0.62

oÄŁ

0.60

its

0.59

our

0.58

Activations Density 0.692%

phrases indicating precision or spatial relationships such as "to the north" or "about 30 kilometers"

prepositions and phrases indicating direction or position

No Comments

No Known Activations