INDEX

Explanations

instances where someone's job, living conditions, or personal life are described in detail

oai_token-act-pair · gpt-3.5-turbo

repeated uses of the word "was."

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

Not in Any Lists

No Comments

Negative Logits

Ê

-0.64

cair

-0.64

 Applicant

-0.62

 Pair

-0.60

inctions

-0.60

gered

-0.58

thora

-0.56

 thous

-0.56

 Sapp

-0.55

ulic

-0.54

POSITIVE LOGITS

ivas

0.74

senal

0.67

esthetic

0.66

anium

0.65

advertising

0.62

 pawn

0.62

been

0.60

Morning

0.59

likely

0.58

oute

0.56

Activations Density 0.250%

instances where someone's job, living conditions, or personal life are described in detail

repeated uses of the word "was."

No Comments

No Known Activations