INDEX

Explanations

mentions of the word "dozens"

oai_token-act-pair · gpt-3.5-turbo

references to a quantity of people or items, specifically the word "dozens" and similar terms

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hegemony

-0.68

Dub

-0.62

ismo

-0.61

abad

-0.61

agate

-0.61

MER

-0.60

 Lover

-0.59

Forge

-0.59

ffen

-0.59

sole

-0.58

POSITIVE LOGITS

entimes

1.01

uously

0.91

eenth

0.85

 ILCS

0.82

 thousand

0.82

uous

0.81

ensical

0.80

ourcing

0.80

hovah

0.78

irtual

0.75

Activations Density 0.018%

mentions of the word "dozens"

references to a quantity of people or items, specifically the word "dozens" and similar terms

No Comments

No Known Activations