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

ater

-0.84

bsite

-0.73

 funnel

-0.73

ides

-0.68

ject

-0.68

ading

-0.67

usters

-0.67

uster

-0.67

ide

-0.66

rel

-0.65

POSITIVE LOGITS

 Frem

0.81

0.80

ãĤ¼ãĤ¦ãĤ¹

0.80

DragonMagazine

0.77

 Cabin

0.76

Wyr

0.74

Osc

0.73

Lum

0.71

 redes

0.71

 glim

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.