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

dash

-0.71

 Niet

-0.70

yip

-0.70

vana

-0.70

igible

-0.69

ument

-0.68

ancial

-0.66

INGTON

-0.63

gotten

-0.63

arten

-0.62

POSITIVE LOGITS

IUM

0.72

ica

0.70

Eye

0.67

Prev

0.65

arov

0.64

iques

0.63

OFF

0.62

rine

0.62

Fra

0.61

agra

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.