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

claw

-0.88

ERA

-0.78

è¦ļéĨĴ

-0.77

ONY

-0.76

olate

-0.72

ORN

-0.70

Sphere

-0.70

ress

-0.70

cffffcc

-0.69

POSITIVE LOGITS

 congestion

0.66

 Fernand

0.66

mercial

0.65

 extrad

0.65

 aesthetic

0.63

 indexed

0.63

 inhibited

0.62

 entertained

0.61

ierrez

0.61

 divers

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.