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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pta

-0.69

hid

-0.68

 dashed

-0.61

bish

-0.60

 sarc

-0.60

je

-0.59

 reserved

-0.58

yk

-0.57

 Soldier

-0.57

 bunk

-0.57

POSITIVE LOGITS

"],"

0.70

ilaterally

0.70

abol

0.67

 Recall

0.66

ugu

0.65

orting

0.64

meet

0.64

CVE

0.62

ALP

0.62

HER

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.