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

ŃĶ

-0.94

 reluct

-0.73

"]=>

-0.73

 princ

-0.70

dimension

-0.66

ailability

-0.65

ACTION

-0.64

sche

-0.62

cffffcc

-0.62

catentry

-0.62

POSITIVE LOGITS

Zed

0.76

vez

0.72

ards

0.70

 Castro

0.67

 Miranda

0.64

 Calder

0.63

abis

0.63

 Mastery

0.62

HS

0.62

arding

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.