INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions or commands

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

 Entered

-0.76

professional

-0.49

 Arbit

-0.48

>[

-0.47

 Hogan

-0.46

 Student

-0.46

-0.45

 Judicial

-0.45

>:

-0.45

iliar

-0.44

POSITIVE LOGITS

wings

0.63

mbuds

0.60

è¦ļéĨĴ

0.60

panic

0.59

Ń·

0.58

insky

0.57

¶æ

0.56

fw

0.56

ansk

0.54

 itself

0.54

Activations Density 1.222%

proper nouns

phrases indicating actions or commands

No Comments

No Known Activations