INDEX

Explanations

words or phrases related to adherence, be it to rules, principles, advice, or universal laws

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ad" used 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Carbuncle

-0.70

 Nemesis

-0.70

FUL

-0.68

Jr

-0.64

 Caribbean

-0.62

 Kira

-0.62

 Keeper

-0.61

uckle

-0.57

 Presbyter

-0.57

 Desire

-0.57

POSITIVE LOGITS

rift

1.36

hesion

1.35

hoc

1.34

roit

1.32

elaide

1.32

idas

1.29

ieu

1.23

missible

1.21

oration

1.19

obe

1.16

Activations Density 0.021%

words or phrases related to adherence, be it to rules, principles, advice, or universal laws

instances of the word "ad" used in various contexts

No Comments

No Known Activations