INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ull

-0.83

ioned

-0.78

yss

-0.77

esi

-0.76

eni

-0.75

aughs

-0.74

kr

-0.73

zl

-0.72

ew

-0.72

itte

-0.69

POSITIVE LOGITS

BALL

0.68

 respectively

0.67

 EVER

0.61

 Felix

0.61

 toggle

0.61

Buk

0.61

senal

0.60

 spam

0.59

 Pengu

0.59

 contiguous

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.