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

Negative Logits

 Amon

-0.72

 earnest

-0.66

 Huntington

-0.62

wana

-0.60

 prayer

-0.60

 Nath

-0.59

 heaven

-0.59

 upkeep

-0.58

 Tehran

-0.58

 Spotlight

-0.57

POSITIVE LOGITS

ãĥ¢

0.81

holm

0.79

ãĥ¼ãĥĨãĤ£

0.79

xual

0.75

meier

0.75

ãĥı

0.74

mble

0.74

susp

0.74

wic

0.73

hent

0.73

Activations Density 0.000%

No Known Activations

This feature has no known activations.