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

vice

-0.70

updated

-0.69

gravity

-0.69

FM

-0.66

 Fant

-0.66

TED

-0.63

Newsletter

-0.61

imens

-0.61

atlantic

-0.61

imen

-0.60

POSITIVE LOGITS

ERY

0.71

 mockery

0.65

 reinvest

0.64

 Prosecut

0.64

 Profit

0.64

rentice

0.63

utherford

0.63

 Holmes

0.61

?!"

0.61

hal

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.