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

©¶æ¥µ

-0.69

 Masquerade

-0.66

GPU

-0.64

angles

-0.64

iHUD

-0.64

²¾

-0.63

Pis

-0.61

 stripes

-0.60

 inference

-0.59

issance

-0.58

POSITIVE LOGITS

ãĥ¤

0.76

 Verge

0.70

ervative

0.69

nesty

0.67

isen

0.67

icer

0.64

iden

0.63

illance

0.62

ebin

0.62

 Prom

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.