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

ĸļ

-0.82

uably

-0.76

secut

-0.73

ethnic

-0.72

uyomi

-0.69

 lash

-0.69

igree

-0.68

aughters

-0.67

alus

-0.67

lia

-0.66

POSITIVE LOGITS

ization

1.20

ize

0.87

 Codex

0.84

isation

0.82

IZ

0.79

 Rebirth

0.71

izing

0.71

DM

0.70

ized

0.70

Gam

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.