INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

essors

-0.73

Ô

-0.69

Ð¼

-0.67

ãģĵ

-0.64

 Izan

-0.64

IGN

-0.64

BAT

-0.63

 ILCS

-0.63

åĮ

-0.62

 lett

-0.61

POSITIVE LOGITS

avage

0.72

 inquest

0.71

 Ambro

0.68

Arg

0.68

quart

0.66

cin

0.66

cest

0.65

icter

0.65

Cooldown

0.65

hal

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.