INDEX

Explanations

phrases indicating surprise or disbelief

oai_token-act-pair · gpt-3.5-turbo

expressions of disbelief or the need for assistance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

•layer 5 - surprise

No Comments

Negative Logits

rongh

-0.74

rane

-0.69

owa

-0.67

heading

-0.62

ighth

-0.62

azel

-0.60

isk

-0.58

bush

-0.57

eport

-0.57

chwitz

-0.57

POSITIVE LOGITS

 anymore

0.65

 âĶľ

0.65

 Louie

0.65

Tex

0.62

 Surprise

0.60

aughs

0.60

uitous

0.60

 Presents

0.59

 Vaugh

0.59

sth

0.58

Activations Density 0.083%

phrases indicating surprise or disbelief

expressions of disbelief or the need for assistance

No Comments

No Known Activations