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

rons

-0.67

tones

-0.65

 compositions

-0.65

 detachment

-0.65

RON

-0.63

rity

-0.62

owder

-0.61

 Pratt

-0.61

 chapters

-0.60

 Poly

-0.59

POSITIVE LOGITS

taboola

0.77

externalActionCode

0.70

etheless

0.69

cms

0.67

othe

0.66

mpeg

0.66

assetsadobe

0.65

weet

0.65

"]=>

0.64

 Cosponsors

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.