INDEX

Explanations

No Explanations Found

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

à¼

-0.76

ļéĨĴ

-0.74

 Terran

-0.69

til

-0.69

hooting

-0.65

ËĪ

-0.62

 scrim

-0.62

 fireplace

-0.61

åĭ

-0.60

 scrimmage

-0.60

POSITIVE LOGITS

leck

0.84

yip

0.76

hyde

0.72

kson

0.71

OHN

0.71

loe

0.70

llah

0.70

anchester

0.70

ablishment

0.69

utherland

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.