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

 entreprene

-0.73

©¶æ¥µ

-0.66

dash

-0.65

amacare

-0.61

Fif

-0.60

NTS

-0.60

 Remix

-0.58

 Splash

-0.58

 derivatives

-0.58

 replacements

-0.57

POSITIVE LOGITS

\(\

0.80

arily

0.79

(\

0.76

rend

0.70

enter

0.70

agnetic

0.70

ahan

0.68

ilitary

0.67

lled

0.65

itÃ©

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.