INDEX

Explanations

numbers related to quantities or measurements

oai_token-act-pair · gpt-3.5-turbo

numerical values and statistics related to population and capacity

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

ayn

-0.73

ersen

-0.72

nington

-0.70

etsk

-0.67

plain

-0.66

kell

-0.65

leys

-0.65

pler

-0.63

arnaev

-0.62

 grate

-0.62

POSITIVE LOGITS

 consecutive

0.81

 million

0.75

 magnitude

0.74

 thousand

0.73

lbs

0.70

%-

0.69

 billion

0.68

 centimeters

0.68

 hundred

0.68

0.67

Activations Density 0.139%

numbers related to quantities or measurements

numerical values and statistics related to population and capacity

No Comments

No Known Activations