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

-0.88

MpServer

-0.79

ãĥ³ãĤ¸

-0.77

AAC

-0.74

 Seym

-0.70

olt

-0.69

alty

-0.69

itsch

-0.68

 è£ıè

-0.68

 Arrows

-0.67

POSITIVE LOGITS

 cloaked

0.78

 chamber

0.66

bug

0.65

 duck

0.65

 hurry

0.64

hes

0.63

 detain

0.62

 shrouded

0.62

venient

0.60

erville

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.