INDEX

Explanations

instances where something is physically installed or put in place

oai_token-act-pair · gpt-3.5-turbo

instances of the word "installed."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

sem

-0.85

 Hague

-0.79

houn

-0.75

emic

-0.72

erred

-0.70

wer

-0.69

ooming

-0.68

ï¸ı

-0.67

berman

-0.67

fax

-0.67

POSITIVE LOGITS

 installed

1.06

 installs

0.97

 installations

0.92

 install

0.88

 installation

0.88

 eleph

0.79

 installing

0.78

rone

0.73

 destro

0.73

uries

0.73

Activations Density 0.016%

instances where something is physically installed or put in place

instances of the word "installed."

No Comments

No Known Activations