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

enger

-0.97

cases

-0.72

dry

-0.71

ropolitan

-0.69

definition

-0.68

lower

-0.68

dimensional

-0.66

minute

-0.66

unction

-0.65

kb

-0.65

POSITIVE LOGITS

é¾įå

0.83

ģ«

0.80

 Patreon

0.79

 è£ıç

0.72

oooo

0.68

bered

0.67

oooooooo

0.66

OOOOOOOO

0.62

 enthusi

0.61

isu

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.