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

GY

-0.67

 Zheng

-0.64

broad

-0.62

Sky

-0.62

Hong

-0.60

 Fahrenheit

-0.60

ukemia

-0.60

 mercury

-0.60

seeking

-0.59

 towards

-0.58

POSITIVE LOGITS

esville

0.78

erity

0.71

 admit

0.64

stood

0.63

olin

0.62

din

0.62

Versions

0.61

ned

0.61

xes

0.61

Session

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.