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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ospons

-0.66

ctions

-0.63

 Places

-0.63

 underestimate

-0.62

ullivan

-0.62

oby

-0.62

haz

-0.60

cav

-0.59

 fear

-0.59

 Grizz

-0.59

POSITIVE LOGITS

iston

0.66

ayette

0.66

rite

0.65

regate

0.63

estial

0.63

sent

0.63

byss

0.63

itial

0.63

Finish

0.62

ittal

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.