INDEX

Explanations

years represented as four digits

oai_token-act-pair · gpt-3.5-turbo

instances of the year 2000

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

atories

-0.72

warts

-0.71

senal

-0.69

venge

-0.67

Else

-0.67

artifacts

-0.66

aste

-0.66

ceiver

-0.64

atory

-0.63

akening

-0.63

POSITIVE LOGITS

mAh

0.84

istar

0.76

visors

0.76

Hz

0.73

rpm

0.72

MHz

0.71

hz

0.71

MHz

0.69

kHz

0.68

å¹

0.68

Activations Density 0.018%

years represented as four digits

instances of the year 2000

No Comments

No Known Activations