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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ebted

-0.72

 ãĤµ

-0.68

istration

-0.65

Hung

-0.62

hander

-0.60

Roaming

-0.60

inging

-0.59

âĨ

-0.58

ARP

-0.57

Yan

-0.57

POSITIVE LOGITS

oor

1.07

sworth

0.79

neum

0.77

arie

0.74

 whim

0.73

ombat

0.73

eway

0.72

iggs

0.72

ĵĺ

0.70

adia

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.