INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hou

-0.83

anmar

-0.76

pend

-0.74

Gru

-0.73

unal

-0.72

outed

-0.72

idth

-0.72

vid

-0.72

evil

-0.71

idates

-0.70

POSITIVE LOGITS

 Flavoring

0.77

 guiActiveUn

0.65

 ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½

0.64

 SOFTWARE

0.64

Ô

0.62

 bricks

0.62

ÃĥÃĤ

0.61

awa

0.61

 indo

0.61

 è£ıè¦ļéĨĴ

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.