INDEX

Explanations

phrases or words indicating agreement or approval

oai_token-act-pair · gpt-3.5-turbo

references to agreements or acknowledgments, often expressed through the word "nod."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inventoryQuantity

-0.79

 tabl

-0.75

OTOS

-0.72

 Tycoon

-0.68

 Hebdo

-0.68

 Islands

-0.65

 Journals

-0.64

à¦

-0.63

ãĥł

-0.62

FFER

-0.62

POSITIVE LOGITS

ules

1.15

nod

1.10

atory

1.05

endum

1.03

ule

1.02

emonic

0.97

emon

0.96

ularity

0.95

uling

0.94

ename

0.92

Activations Density 0.025%

phrases or words indicating agreement or approval

references to agreements or acknowledgments, often expressed through the word "nod."

No Comments

No Known Activations