INDEX

Explanations

listings or mentions of items or information

oai_token-act-pair · gpt-3.5-turbo

instances of the word "listing" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Princ

-0.69

Engineers

-0.67

udge

-0.66

Fu

-0.64

acies

-0.64

iang

-0.62

hei

-0.62

stad

-0.61

aceutical

-0.61

het

-0.59

POSITIVE LOGITS

 listing

3.79

 listings

2.51

 listed

1.62

 list

1.57

 List

1.57

list

1.45

listed

1.39

 lists

1.36

 LIST

1.35

List

1.35

Activations Density 0.013%

listings or mentions of items or information

instances of the word "listing" and its variations

No Comments

No Known Activations