INDEX

Explanations

phrases expressing agreement

oai_token-act-pair · gpt-3.5-turbo

expressions of agreement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ãĤ¼ãĤ¦ãĤ¹

-0.78

akin

-0.76

mAh

-0.71

chin

-0.71

 Dise

-0.69

quer

-0.66

mage

-0.64

cade

-0.64

GV

-0.63

patch

-0.62

POSITIVE LOGITS

 regards

1.34

 regard

1.22

 respect

1.09

standing

0.93

 impunity

0.83

stood

0.78

 destiny

0.74

 sincerity

0.73

ĪĴ

0.71

hus

0.71

Activations Density 0.056%

phrases expressing agreement

expressions of agreement

No Comments

No Known Activations