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

ancial

-0.80

 brill

-0.74

adobe

-0.69

ement

-0.65

 Photograph

-0.65

 Agric

-0.65

avorite

-0.64

abase

-0.64

cellence

-0.63

affer

-0.62

POSITIVE LOGITS

>[

0.75

ourn

0.71

bin

0.70

ãĥı

0.68

éĢ

0.68

èĢħ

0.65

 unborn

0.64

Ã»

0.64

gyn

0.63

oused

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.