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

 unden

-0.88

 contrad

-0.72

 Definitive

-0.71

 Seah

-0.69

 resemb

-0.67

 reluct

-0.66

 conclud

-0.65

 Authent

-0.63

 indic

-0.63

 Presents

-0.62

POSITIVE LOGITS

zo

0.77

mor

0.76

 Denis

0.74

akings

0.69

asons

0.66

gres

0.66

agers

0.65

kins

0.65

-------

0.63

anu

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.