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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lists

-0.70

otes

-0.63

 lips

-0.62

ãĥ£

-0.62

ilk

-0.60

ogie

-0.59

 Niet

-0.59

ateg

-0.59

utes

-0.59

ggie

-0.58

POSITIVE LOGITS

APR

0.71

nikov

0.68

charge

0.68

ahime

0.67

LTD

0.65

adr

0.65

incerity

0.64

inkle

0.64

efficiency

0.64

awa

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.