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

successful

-0.74

akespe

-0.69

soDeliveryDate

-0.68

critical

-0.66

Pic

-0.63

eston

-0.63

 scen

-0.62

inning

-0.62

lication

-0.62

interrupted

-0.62

POSITIVE LOGITS

ãĥ¼ãĥĨ

0.72

ãĥ³ãĤ¸

0.64

enum

0.62

ãĥ¼ãĤ¯

0.60

ãĥ¼

0.60

ESH

0.59

izoph

0.59

ãĥ³

0.59

IRO

0.58

zin

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.