INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

prop

-0.84

Prop

-0.79

osate

-0.77

Assembly

-0.69

otiation

-0.68

Hun

-0.67

activ

-0.66

aten

-0.66

OLOG

-0.65

behind

-0.65

POSITIVE LOGITS

 equally

0.72

 asleep

0.71

 sprinkled

0.70

 governed

0.69

 penetrated

0.67

 amazed

0.67

 neither

0.66

 depended

0.66

 influenced

0.65

 baffled

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.