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

owe

-0.76

uit

-0.76

idas

-0.76

iates

-0.74

ourcing

-0.74

orne

-0.73

conn

-0.73

anguage

-0.71

itiz

-0.71

ornia

-0.70

POSITIVE LOGITS

 Hoover

0.78

 Swordsman

0.69

 Clause

0.64

 Dwar

0.63

 Calculator

0.62

 Heroic

0.62

Sov

0.62

 Controlled

0.62

LAW

0.61

 Pengu

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.