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

akura

-0.70

NF

-0.70

ebted

-0.69

ARC

-0.67

HRC

-0.66

ItemImage

-0.65

 sidx

-0.64

 reacting

-0.64

 actionGroup

-0.63

 Hanson

-0.63

POSITIVE LOGITS

forth

0.78

 jaws

0.76

earch

0.73

bum

0.72

gypt

0.69

ourt

0.69

ntil

0.68

lished

0.68

 surv

0.67

peg

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.