INDEX

Explanations

phrases related to community and unity

oai_token-act-pair · gpt-3.5-turbo

references to categories or classifications and their associated properties

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

ISON

-0.65

iHUD

-0.63

 arrang

-0.62

 itself

-0.61

enhagen

-0.60

jandro

-0.59

>:

-0.58

\\\\\\\\

-0.58

ãĤ´

-0.58

 descriptor

-0.55

POSITIVE LOGITS

romy

0.82

etsk

0.75

 Emer

0.71

lime

0.70

hift

0.68

ocr

0.68

ettings

0.66

hots

0.64

 Confederacy

0.63

peed

0.63

Activations Density 0.935%

phrases related to community and unity

references to categories or classifications and their associated properties

No Comments

No Known Activations