INDEX

Explanations

the word "what" followed by an action verb, representing a call to action or identification of tasks to be performed

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "do what" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Yards

-0.65

 Returning

-0.64

 Vaugh

-0.59

end

-0.58

imming

-0.58

 inacc

-0.58

UTH

-0.57

 Cong

-0.56

uttering

-0.56

ohyd

-0.56

POSITIVE LOGITS

soever

1.09

Downloadha

0.93

 happens

0.85

 else

0.74

 mattered

0.71

ĸļ

0.71

 kinds

0.71

 happened

0.70

odynam

0.69

 inspires

0.67

Activations Density 0.097%

the word "what" followed by an action verb, representing a call to action or identification of tasks to be performed

instances of the phrase "do what" in various contexts

No Comments

No Known Activations