INDEX

Explanations

prepositions indicating location or direction

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a notable status or rank among multiple entities

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

getic

-0.65

zeb

-0.63

 lapt

-0.60

gee

-0.59

 prol

-0.58

iture

-0.57

culosis

-0.55

hya

-0.54

desktop

-0.53

 Provided

-0.53

POSITIVE LOGITS

those

0.72

 those

0.72

 many

0.70

few

0.66

apest

0.65

few

0.64

icial

0.64

ãĤ¨

0.64

arching

0.64

abet

0.63

Activations Density 0.073%

prepositions indicating location or direction

phrases indicating a notable status or rank among multiple entities

No Comments

No Known Activations