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

lihood

-0.72

 Doodle

-0.69

yrs

-0.63

 stump

-0.62

claimer

-0.62

hent

-0.61

 Rochester

-0.61

estones

-0.61

athon

-0.60

 Roose

-0.58

POSITIVE LOGITS

0.81

oldown

0.76

ibur

0.74

rade

0.73

ractions

0.73

CLASSIFIED

0.70

icipated

0.70

interstitial

0.69

 Dres

0.69

linger

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.