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

 subscribed

-0.70

 panicked

-0.66

doors

-0.64

imar

-0.63

uran

-0.62

ibur

-0.61

arya

-0.61

 understatement

-0.61

 patience

-0.60

 illiter

-0.60

POSITIVE LOGITS

ãĥķãĤ©

0.83

 æľ

0.78

ocamp

0.78

çĭ

0.76

å§

0.75

phen

0.75

ãĥĳ

0.74

GSL

0.70

 Pengu

0.70

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.