INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

udic

-0.75

itability

-0.69

DragonMagazine

-0.67

utterstock

-0.65

 technologically

-0.65

 entertain

-0.63

 spectators

-0.63

riks

-0.63

azeera

-0.63

uploads

-0.62

POSITIVE LOGITS

CCC

0.82

ãĤ¡

0.82

EMS

0.79

KA

0.78

Ò

0.77

APS

0.77

LAB

0.76

ECA

0.76

ãĥ¼ãĥ³

0.74

Xi

0.74

Activations Density 0.000%

No Known Activations

This feature has no known activations.