INDEX

Explanations

references to something being alive or potentially existing

oai_token-act-pair · gpt-3.5-turbo

references to being alive or in a state of existence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

RECT

-0.69

é¾įå

-0.67

ãĥĩ

-0.64

UGC

-0.59

ciples

-0.59

 Telescope

-0.58

awaru

-0.58

agree

-0.58

FG

-0.57

ims

-0.57

POSITIVE LOGITS

lihood

1.10

abouts

0.87

lier

0.82

weight

0.81

spin

0.75

 alive

0.74

nces

0.72

terday

0.72

mares

0.70

beat

0.68

Activations Density 0.043%

references to something being alive or potentially existing

references to being alive or in a state of existence

No Comments

No Known Activations