INDEX

Explanations

references to a specific person named "Lor"

oai_token-act-pair · gpt-3.5-turbo

the word "lor" and its variations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Governments

-0.71

BOOK

-0.67

 Somali

-0.65

 Unch

-0.65

ÂłÂłÂł

-0.64

ÂłÂł

-0.64

 Advocate

-0.63

OPLE

-0.63

tails

-0.63

 Eliot

-0.62

POSITIVE LOGITS

iculture

1.10

ion

0.98

lor

0.97

inated

0.92

ite

0.92

ith

0.90

oso

0.89

icultural

0.89

apy

0.88

ious

0.88

Activations Density 0.008%

references to a specific person named "Lor"

the word "lor" and its variations

No Comments

No Known Activations