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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

needs

-0.72

 Shutdown

-0.68

 overfl

-0.63

 Nights

-0.63

uren

-0.63

Kab

-0.61

lez

-0.61

had

-0.60

 Delta

-0.59

 Tournament

-0.58

POSITIVE LOGITS

ierre

0.83

aughs

0.82

 then

0.80

catentry

0.79

ARDIS

0.79

untary

0.72

apon

0.71

itably

0.71

uve

0.70

 THEN

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.