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

ormons

-0.75

undo

-0.73

ription

-0.72

RIC

-0.70

regor

-0.66

paces

-0.64

rams

-0.64

DERR

-0.62

 tremend

-0.62

 oats

-0.61

POSITIVE LOGITS

0.70

 Hasan

0.68

0.67

00200000

0.62

 Sultan

0.61

measures

0.61

0.59

ifling

0.59

 Hear

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.