INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ãĤ¶

-0.77

 redes

-0.71

webkit

-0.71

ahime

-0.70

ococ

-0.68

bay

-0.68

etsk

-0.67

 looph

-0.66

minecraft

-0.64

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.64

POSITIVE LOGITS

INGS

0.77

 declining

0.66

 yours

0.65

 declines

0.64

 dwindling

0.64

erence

0.63

hes

0.62

 sober

0.60

erent

0.60

 Honor

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.