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

qua

-0.68

Boe

-0.67

bered

-0.64

 Magazine

-0.64

 Nights

-0.63

 Illustrated

-0.63

Faw

-0.62

bies

-0.62

 Babe

-0.62

 Hath

-0.61

POSITIVE LOGITS

cible

0.73

Downloadha

0.71

ãĤ¬

0.70

semble

0.66

ichen

0.66

 cryptoc

0.65

Ground

0.64

vol

0.63

ãĤ®

0.63

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.