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

idable

-0.83

irlwind

-0.75

puter

-0.72

utherford

-0.69

sburgh

-0.68

swer

-0.68

ortality

-0.67

alysis

-0.65

Äĩ

-0.65

riages

-0.65

POSITIVE LOGITS

çī

0.72

 lawy

0.72

 substit

0.67

 guiActiveUn

0.66

 subtitles

0.66

uca

0.63

oti

0.63

 transc

0.60

èĢħ

0.58

èĥ

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.