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

Roaming

-0.76

SourceFile

-0.67

 rune

-0.63

emetery

-0.62

Tile

-0.62

Num

-0.62

 srfAttach

-0.62

Clear

-0.60

Probably

-0.60

ricia

-0.60

POSITIVE LOGITS

 Spears

0.81

 Buckley

0.75

 Coulter

0.71

 Livingston

0.69

atform

0.69

EW

0.67

 Beir

0.66

ection

0.64

verage

0.62

 VIDEOS

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.