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

IVES

-0.86

reprene

-0.80

ĵĺ

-0.79

ives

-0.77

aneers

-0.77

redo

-0.75

abet

-0.73

urious

-0.72

URES

-0.71

 trave

-0.71

POSITIVE LOGITS

 Maid

0.77

terson

0.69

 Amon

0.68

mor

0.67

 Eighth

0.66

otin

0.66

 Feder

0.65

 Lutheran

0.65

 Gillespie

0.65

 Shroud

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.