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

lyak

-0.82

arching

-0.68

aez

-0.68

[&

-0.65

ansk

-0.62

Kus

-0.62

atis

-0.61

 Krug

-0.61

gew

-0.60

baugh

-0.59

POSITIVE LOGITS

Cond

0.77

Merc

0.73

é¾

0.70

IER

0.69

Mp

0.68

ORD

0.68

 Compass

0.67

rg

0.66

FACE

0.66

ENGTH

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.