INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

cair

-0.93

Downloadha

-0.88

byss

-0.74

heast

-0.73

seys

-0.73

GROUND

-0.73

ailability

-0.72

PDATE

-0.72

 Ambro

-0.70

¦

-0.70

POSITIVE LOGITS

iques

0.73

 stub

0.65

avis

0.64

anos

0.64

 brig

0.63

 ruins

0.62

od

0.61

abad

0.61

akia

0.60

Rog

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.