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

owed

-0.71

ier

-0.65

":[

-0.65

tru

-0.65

-+-+-+-+

-0.65

 Edison

-0.64

rn

-0.62

 Cipher

-0.60

 attendance

-0.60

Ry

-0.59

POSITIVE LOGITS

poke

0.93

 Doodle

0.83

TAG

0.83

İĭ

0.82

gypt

0.74

Palest

0.71

pees

0.71

 Cosponsors

0.70

yout

0.67

ternity

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.