INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 conclud

-0.78

ãĤ¦ãĤ¹

-0.72

 Bake

-0.70

 Appeal

-0.69

EY

-0.67

Reviewer

-0.67

ften

-0.66

 ãĤµ

-0.64

bright

-0.64

istically

-0.63

POSITIVE LOGITS

 bears

0.74

 unloaded

0.70

 conver

0.68

ocr

0.66

ru

0.65

rou

0.62

 floats

0.60

 pipelines

0.60

owing

0.60

 encaps

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.