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

 Nicarag

-0.72

ospace

-0.71

connected

-0.64

 Pens

-0.63

 Airport

-0.63

 contrace

-0.59

Virtual

-0.59

Nex

-0.59

ccess

-0.58

equipped

-0.58

POSITIVE LOGITS

gow

0.82

enge

0.76

forth

0.75

ingham

0.69

shaw

0.68

 hammer

0.67

bil

0.67

ual

0.65

sag

0.65

bow

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.