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

 Parenthood

-0.79

 preaching

-0.73

TPPStreamerBot

-0.69

ħĭ

-0.69

luaj

-0.69

 Malays

-0.68

cult

-0.67

phas

-0.65

Downloadha

-0.62

 estrogen

-0.62

POSITIVE LOGITS

 Luxem

0.67

reement

0.67

oided

0.67

encies

0.67

uez

0.66

rics

0.66

 envis

0.65

ortium

0.64

elt

0.63

 Greenwald

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.