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

ides

-0.74

resident

-0.66

iding

-0.64

actively

-0.63

eteenth

-0.63

oan

-0.63

 pans

-0.61

pan

-0.61

opes

-0.61

anders

-0.61

POSITIVE LOGITS

DragonMagazine

0.77

 Computing

0.74

catentry

0.74

abal

0.73

ãĥĨãĤ£

0.70

PCR

0.68

Cth

0.68

Memory

0.67

INGTON

0.66

#$

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.