INDEX

Explanations

proper nouns related to people's names

oai_token-act-pair · gpt-3.5-turbo

proper names, particularly surnames

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

 citiz

-0.76

soDeliveryDate

-0.74

imeters

-0.71

merga

-0.70

ntil

-0.69

udeb

-0.69

ultan

-0.68

 conflic

-0.68

GoldMagikarp

-0.68

omed

-0.67

POSITIVE LOGITS

 County

0.99

ville

0.95

Jr

0.88

stown

0.88

 Township

0.87

AFB

0.82

 Parish

0.79

's

0.79

Collins

0.79

 Goodman

0.78

Activations Density 0.249%

proper nouns related to people's names

proper names, particularly surnames

No Comments

No Known Activations