INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Ogre

-0.70

+/-

-0.68

 awoken

-0.67

OG

-0.67

 margins

-0.66

oshenko

-0.65

 Trace

-0.63

 porous

-0.63

 Abstract

-0.62

 Split

-0.62

POSITIVE LOGITS

xual

0.91

sha

0.69

 Administrator

0.65

ãĥĸ

0.64

 lesbian

0.62

ctl

0.61

igate

0.61

 moth

0.61

 Regulatory

0.60

minist

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.