INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

hes

-0.68

iates

-0.67

 pneum

-0.65

lic

-0.64

 yeast

-0.64

 neglig

-0.64

 projectiles

-0.64

 introdu

-0.63

 adults

-0.63

iated

-0.63

POSITIVE LOGITS

ä»£

0.79

ETF

0.79

LAND

0.79

Court

0.74

 Watergate

0.72

ANA

0.72

iverpool

0.71

sonian

0.71

 GOODMAN

0.71

 subpoena

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.