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

anwhile

-0.78

Thumbnail

-0.69

 Sidd

-0.69

EVA

-0.69

 Alchemist

-0.66

 berth

-0.65

anship

-0.64

cules

-0.64

avorite

-0.63

 personalities

-0.63

POSITIVE LOGITS

Fax

0.73

rors

0.69

amsung

0.66

 Harbor

0.66

Win

0.64

her

0.63

grim

0.62

ocker

0.62

 Skydragon

0.60

rag

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.