INDEX

Explanations

phrases related to specific places in Europe

oai_token-act-pair · gpt-3.5-turbo

references to specific geographical locations or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

atorium

-0.80

hov

-0.79

 Bolshe

-0.78

 McGill

-0.75

 persecut

-0.72

 USSR

-0.72

 labour

-0.69

 persecution

-0.67

 persuasion

-0.67

 exhibition

-0.67

POSITIVE LOGITS

 warrants

1.33

ppa

1.25

 Patch

1.20

 Seas

1.12

leaf

1.12

Jol

1.10

WB

1.08

Patch

1.07

 Legendary

1.05

 Parkway

1.00

Activations Density 0.053%

phrases related to specific places in Europe

references to specific geographical locations or entities

No Comments

No Known Activations