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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 eleph

-0.79

atem

-0.72

gart

-0.72

eger

-0.70

obiles

-0.68

 withd

-0.68

acci

-0.67

MQ

-0.67

JM

-0.67

avier

-0.67

POSITIVE LOGITS

 Pastebin

0.80

 Horde

0.69

 writ

0.67

 Trojan

0.64

 Founders

0.63

 Trident

0.62

 Bungie

0.61

 Legislature

0.61

 plurality

0.60

 GEAR

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.