INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

bet

-0.71

thy

-0.69

plot

-0.68

card

-0.63

peak

-0.62

 afore

-0.62

Ù

-0.61

burgh

-0.61

Draft

-0.61

rete

-0.60

POSITIVE LOGITS

inar

0.76

essee

0.74

soDeliveryDate

0.71

::::::::

0.70

epad

0.69

issance

0.68

osaurus

0.67

abilia

0.67

yna

0.67

acea

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.