INDEX

Explanations

instances of the term "Em" followed by a number indicating intensity

oai_token-act-pair · gpt-3.5-turbo

references to the name "Emory."

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

 Kubrick

-0.83

 Icelandic

-0.75

 Elves

-0.71

 Iceland

-0.69

 Idle

-0.69

ãģį

-0.69

 sund

-0.67

lihood

-0.67

WAYS

-0.67

ASC

-0.65

POSITIVE LOGITS

phasis

1.33

otions

1.32

otional

1.32

igrant

1.26

otion

1.23

igrants

1.19

brace

1.18

blems

1.15

ission

1.15

perors

1.14

Activations Density 0.015%

instances of the term "Em" followed by a number indicating intensity

references to the name "Emory."

No Comments

No Known Activations