INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 blasphemy

-0.72

Eva

-0.71

ById

-0.65

 negativity

-0.65

Altern

-0.63

SourceFile

-0.62

¬¼

-0.61

NB

-0.61

 Partners

-0.60

Shares

-0.60

POSITIVE LOGITS

 scroll

0.83

nels

0.75

groups

0.74

 TOTAL

0.74

agons

0.69

ults

0.69

okane

0.69

 guiActiveUnfocused

0.68

 ISBN

0.68

catentry

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.