INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 protector

-0.77

 Samurai

-0.69

 liner

-0.67

 initials

-0.66

ARDIS

-0.65

ain

-0.65

good

-0.64

_-

-0.63

 Petroleum

-0.62

 guardian

-0.61

POSITIVE LOGITS

getic

0.77

bors

0.76

»

0.70

©¶æ¥µ

0.69

ezvous

0.69

©¶æ

0.69

 Procedure

0.68

forcement

0.68

initions

0.68

 Cantor

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.