INDEX

Explanations

specific product names, locations, and terms associated with technology and media products

oai_token-act-pair · gpt-3.5-turbo

references to media, technology, or materials

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

 renamed

-0.59

guyen

-0.57

anwhile

-0.54

00200000

-0.51

 Variant

-0.49

quartered

-0.48

 consists

-0.47

uilt

-0.47

ropolitan

-0.46

surprisingly

-0.46

POSITIVE LOGITS

.ï¿½

0.56

.''

0.55

.</

0.53

$.

0.48

".

0.47

?".

0.46

versive

0.46

.}

0.44

 NEXT

0.43

Pax

0.42

Activations Density 2.786%

specific product names, locations, and terms associated with technology and media products

references to media, technology, or materials

No Comments

No Known Activations