INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

oops

-0.79

ucl

-0.71

owsky

-0.70

cedented

-0.66

ropri

-0.65

acts

-0.63

ova

-0.63

 Dragonbound

-0.63

urate

-0.62

andro

-0.61

POSITIVE LOGITS

kees

0.89

zos

0.72

atre

0.72

 plantation

0.66

 ______

0.65

 stump

0.64

ILLE

0.63

 Timbers

0.63

 Territ

0.62

 Chimera

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.