INDEX

Explanations

numerical ranges related to quantities or measurements

oai_token-act-pair · gpt-3.5-turbo

numerical values, particularly those indicating quantities related to demographics, statistics, or financial data

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

Reviewer

-0.70

Viz

-0.66

 Drawn

-0.63

 Nationwide

-0.61

 Doodle

-0.59

went

-0.59

tery

-0.59

Tus

-0.58

atem

-0.56

 Prelude

-0.56

POSITIVE LOGITS

mph

0.77

ulia

0.74

 feet

0.70

MPG

0.68

mph

0.67

morrow

0.66

kHz

0.64

uate

0.64

rouse

0.63

amber

0.62

Activations Density 0.114%

numerical ranges related to quantities or measurements

numerical values, particularly those indicating quantities related to demographics, statistics, or financial data

No Comments

No Known Activations