INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 quir

-0.77

dab

-0.70

terday

-0.68

 aided

-0.66

 participates

-0.66

fully

-0.66

 behalf

-0.63

 Arabia

-0.63

 malf

-0.63

 airplanes

-0.61

POSITIVE LOGITS

anie

0.76

kees

0.75

 Sims

0.72

ocial

0.69

img

0.67

asma

0.65

ondo

0.65

arte

0.64

 Roose

0.63

uten

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.