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

esan

-0.75

Eag

-0.74

 glim

-0.68

%]

-0.66

Ø¹

-0.66

Ø¯

-0.65

 Dunham

-0.65

 Shack

-0.65

 blat

-0.65

Ø³

-0.64

POSITIVE LOGITS

ilit

0.77

razil

0.72

upiter

0.70

psey

0.69

insk

0.69

ocket

0.67

ueller

0.67

aeus

0.66

leck

0.65

irted

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.