INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ories

-0.79

 roofs

-0.73

ury

-0.73

xual

-0.72

 solder

-0.71

scribe

-0.71

perse

-0.70

ocene

-0.68

lde

-0.68

izabeth

-0.68

POSITIVE LOGITS

 Carrier

0.76

 Misc

0.72

Incre

0.65

calling

0.65

errors

0.62

pg

0.62

misc

0.59

obook

0.59

percent

0.58

epad

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.