INDEX

Explanations

the term "exhibit."

oai_token-act-pair · gpt-4-turbo

references to art exhibitions or displays

oai_token-act-pair · gpt-3.5-turbo

references to exhibitions or displays

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

srf

-0.71

 payday

-0.70

etheless

-0.68

pop

-0.67

apo

-0.67

leans

-0.64

 waged

-0.64

nington

-0.63

sembly

-0.63

 moons

-0.63

POSITIVE LOGITS

 exhibits

1.32

 exhibit

1.11

ĸļ

1.04

 displays

0.96

 Exhibit

0.94

 exhibitions

0.89

 exhibited

0.87

erers

0.84

emonium

0.81

 showcases

0.80

Activations Density 0.014%

the term "exhibit."

references to art exhibitions or displays

references to exhibitions or displays

No Comments

No Known Activations