INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ARP

-0.82

IRO

-0.72

 ashes

-0.71

 Arrow

-0.70

Gre

-0.69

 Reload

-0.69

 Shards

-0.66

 Gems

-0.65

ANG

-0.65

 bindings

-0.65

POSITIVE LOGITS

abouts

0.83

formation

0.79

cision

0.77

tle

0.75

come

0.74

intestinal

0.73

street

0.73

berman

0.70

moral

0.70

seek

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.