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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asc

-0.73

hing

-0.67

 propag

-0.65

 prem

-0.63

thouse

-0.62

UTC

-0.62

 architect

-0.62

raper

-0.61

anke

-0.60

emouth

-0.60

POSITIVE LOGITS

BILL

0.76

bably

0.74

 bruises

0.72

ãĤ½

0.71

 Violence

0.70

 sovere

0.70

 Innocent

0.68

EVA

0.67

aneously

0.67

duty

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.