INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

clusion

-0.67

arse

-0.63

isans

-0.62

 sunset

-0.61

 streak

-0.60

join

-0.60

ilater

-0.59

 horse

-0.59

 frontier

-0.58

 fringe

-0.58

POSITIVE LOGITS

ugu

0.75

 reckoned

0.67

uke

0.67

 Understand

0.66

semb

0.64

annabin

0.64

bourg

0.64

channelAvailability

0.63

 sake

0.63

hiba

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.