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

myra

-0.82

Ly

-0.74

rified

-0.70

ocobo

-0.69

 McGr

-0.69

ipl

-0.68

rique

-0.67

 Weber

-0.67

 Norris

-0.67

tg

-0.67

POSITIVE LOGITS

who

0.88

who

0.82

 whom

0.81

river

0.72

Assembly

0.68

hest

0.67

åŃĲ

0.66

whose

0.64

father

0.64

ãĤ¼ãĤ¦ãĤ¹

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.