INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ploy

-0.92

ocations

-0.76

 Dull

-0.69

igator

-0.69

irs

-0.68

udeb

-0.67

othe

-0.65

pas

-0.65

ocate

-0.63

ocation

-0.63

POSITIVE LOGITS

qus

0.76

 adolesc

0.72

ãĤ¦ãĤ¹

0.71

 turnout

0.67

çīĪ

0.66

ãĤ¨

0.65

 footing

0.64

ä¸Ģ

0.64

 tide

0.62

ãĤ¤ãĥĪ

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.