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

itized

-0.76

hered

-0.74

.")

-0.66

 è£ıè

-0.65

adin

-0.64

lot

-0.64

DAY

-0.62

Wilson

-0.62

FN

-0.61

 compress

-0.61

POSITIVE LOGITS

Policy

0.75

ipment

0.73

apons

0.69

obbies

0.67

 backdrop

0.66

inki

0.64

 embargo

0.62

utenberg

0.61

 Subcommittee

0.60

Director

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.